Vous êtes sur la page 1sur 508

Quantum Mechanics,

Mathematics,
Cognition and Action

Proposals for a Formalized Epistemology

Edited by

Mioara Mugur-Schächter

Centre pour la Synthèse d’une Épistemologie Formalisée


Paris, France

and

Alwyn van der Merwe

University of Denver
Denver, Colorado, U.S.A.

KLUWER ACADEMIC PUBLISHERS


DORDRECHT/BOSTON/LONDON
CONTENTS

Preface vii

Introduction ix

Part One: Preliminary Explorations: What, Why, How? 1

1. Remarks About the Program for a Formalized Epistemology


Francis Bailly 3
2. Formalized Epistemology in a Philosophical Perspective
Hervé Barreau 9
3. Formalized Epistemology, Logic, and Grammar
Michel Bitbol 21
4. Epistemic Operations and Formalized Epistemology:
Contribution to the Study of the Role of Epistemic
Operations in Scientific Theories
Michel Paty 37
5. Mathematical Physics and Formalized Epistemology:
Debate with Jean Petitot
Interlocutors: Francis Bailly, Michel Bitbol,
Mioara Mugur-Schächter, Vincent Schächter 73
6. On the Possibility of a Formalized Epistemology
Robert Valleée 103

Part Two: Constructive Contributions 107

7. Quantum Mechanics Versus a Method of Relativized


Conceptualization
Mioara Mugur-Schächter 109
8. Mathematical and Formalized Epistemologies
Robert Vallée 309
9. Ago-Antagonistic Systems
Élie Bernard-Weil 325

v
vi

Part Three: Further Explorations 349

10. Complexity of the “Basic Unit” of Language:


Some Parallels in Physics and Biology
Evelyne Andreewsky 351
11. About the Emergence of Invariances in Physics:
from “Substantial” Conservation to Formal Invariance
Francis Bailly 369
12. Form and Actuality
Michel Bitbol 389
13. To Suspended Informal Time
Michel Paty 431
14. The Constructed Objectivity of the Mathematics
and the Cognitive Subject
Giuseppe Longo 433
15. On Complexity
Vincent Schächter 463

Appendix: Biographical Notes 487

Author and Subject Index 491


PREFACE

The Centre pour la Synthèse d’une Épistemologie Formalisée, henceforth


briefly named CeSEF, was founded in June 1994 by a small group of sci-
entists working in various disciplines, with the definite aim to synthesize a
“formalized epistemology” founded on the methods identifiable within the
foremost modern scientific disciplines. Most of the founders were already
authors of well-known works displaying a particular sensitivity to epistemo-
logical questions. But the aim that united us was new. This aim along with
the peculiar choice of its verbal expression are thoroughly discussed in the
Introduction.
In the present volume, we publish the first harvest of explorations and
constructive proposals advanced in pursuit of our goal. The contributions
are expressive also of the views of those who shared only our beginnings and
then left us1 ; they equally reflect input from those who participated in our
workshops but did not contribute to this volume.
We are indebted to the Association Naturalia et Biologica for having
supported with a donation the publication of this volume.
The camera-ready form of this book we owe to the patient and metic-
ulous labor of Ms. Jackie Gratrix. The superb job she has done is herewith
gratefully acknowledged.

Mioara Mugur-Schächter and Alwyn van der Merwe

1
Paul Bourgine and, quite specially, Bernard Walliser.

vii
viii
INTRODUCTION

The purpose of this book is to initiate a new discipline, namely a formalized


epistemological method drawn both from the cognitive strategies practised
in the main modern sciences and from general philosophical thinking. Our
progress in this direction will be attempted by general discussions concern-
ing the concept itself, by constructive attempts, and by informative-critical
explorations. Our goal has been triggered by the following considerations.
Everywhere at the present frontiers of scientific thought one can
watch how absolute assertions and absolute separations that formerly
seemed unshakeable are fading away.
So, for instance, in logic and mathematics the belief in the possibility
of an uninterrupted progression of unlimited purely formal developments,
which dominated the beginning of the last century, has collapsed. It has be-
come clear that any definite domain of exclusively formal action is confined,
even if in principle it can always be extended, while the process of extension
itself escapes formalization, as also, quite radically, the process of creation
of a domain of formal operationality does.
For living systems, the definition of what is called the system raises
nontrivial problems. Biologists have been led to introduce notions like “self-
organization” and “organizational closure” in order to point the way in which
a living system constantly re-constructs its own matter, forms, and functions
by processes where the feedback upon the system, of its interactions with
the environment, are as important as the characters of the system itself.
As soon as life is involved, the concept of cause resists any attempt
to clearly distinguish it from the concept of aim. For living beings as well as
for those meta-living beings called social organizations, the importance of
pragmatic models conforming to aims located in the future but shaping the
actions accomplished now in order to reach the aims, becomes decisive. The
aims—tied to belief and anticipation—operate backwards upon the action
that furthers the aims, whereas the action, while it develops, changes the
aims. This entails a dynamic that depends upon its history and its context,
and of which the characterization requires a cognitivistic and evolutionary
approach.
The theory of (the communication of) information deals with the
transmission of messages by making use of a probabilistic representation of a
peculiar sort, according to which any received message unavoidably depends
not exclusively on the message sent but also on the “channel” used in the

ix
x

process. Thus the message received is quite fundamentally dependent on


the way in which it becomes perceptible to the receiver. As a consequence,
the possiblity of reconstructing the sent message out of the received one has
to be studied explicitly as a function of the modalities of transmission; and
the conditions required for such a reconstruction are highly nontrivial.
The investigators of “chaos” have resolved a millenary confusion by
elaborating abstract mathematical examples, on the one hand, and simu-
lations, on the other hand, which prove that determinism does not entail
predictability: Deterministic modelings, and the full recognition of the ran-
domness of the facts such as they are directly perceived by us exist side by
side in mutual independence. Thus the fictitious belief that a choice has to
be made evaporates, and a world of new questions arises concerning a per-
tinent representation of the relations between perceptual randomness and
deterministic models of physical processes.
In the approaches concerning the treatment of “complex” systems or
processes, the “agents”, their “environment” and “actions”, and the feed-
backs from these, constitute inextricably entwined hierarchies of matter,
situations, conscious aims and behaviours, knowledges, social organizations,
and devices. What is named how, what is treated how, becomes a mat-
ter of method much more than a matter of fact. The boundaries between
categories with fixed inner content fade away, and roles take their place.
And so on. We could continue the list. Everywhere the contours
of separations that seemed obvious, clear-cut and absolute become shaky
and full of gaps. And these superficial symptoms make us feel that we are
witnessing changes which, though superficially appearing to be unrelated,
are connected beneath the level of the directly perceptible. We also feel that
the implications of these changes go down very deep, that they touch and
modify the slopes of the first layer of our conceptualization, the very place
where the general structure of our modern way of thinking and speaking has
been forged. But the nature of changes of this sort—precisely because they
concern established manners of thinking and speaking—is very difficult to
grasp by use of the established manner of thinking and speaking. So the
existence of these changes is revealed by their effects long before we become
able to discern and express their precise content.
The very existence of these changes as such, before any attempt to
define their contents, already raises questions. The conceptualization by
man, of what he calls “reality”, is itself an element of “reality”. Is it then
not subjected to some laws, to some invariances? This should be the case
in some sense; but in which sense exactly? What changes and what stays
the same? How could one delve deep enough, and how should we proceed in
order to be sure that we capture and fully seize the essence of the develop-
INTRODUCTION xi

ing transmutation as well as the stable structure that meanwhile persists?


Without permitting decades to pass while the process is accomplished im-
plicitely by osmotic assimilation of random, disparate bits of knowledge and
interactions among them, without generating any perceivable contour?
It would be of crucial utility to succeed. Only what is explicitly known
acquires a definite form, perceptible from the “outside”. And only once this
happens does it become possible to then detach what has been formed,
optimize it with respect to definite purposes, and shape it into a genuine
instrument that can be deliberately employed and indefinitely improved.
At the beginning of the last century, the theory of special relativity
reduced the structure of the concept of spacetime that underlies the descrip-
tions of physical phenomena, in the sense that the fracture of a bone is re-
duced by a surgeon. And later, starting in 1924, quantum mechanics crafted
conceptual-operational-formal channels that have enabled the human mind
to apply itself directly to the unobservable and to construct concerning it
observable predictions that are realized with impressive precision. Of course,
these are arcane revolutions which so far have penetrated the thinking of
only a very few people. Moreover, they are as yet unfinished revolutions.
But some philosophers, helped by a small number of physicists, have gener-
ated a process of communication by which, osmotically, the essence of some
views of modern physics has more or less infused many minds. The germs of
new approaches that are developing in various areas of scientific investiga-
tion have sprouted in this modified earth, which has nourished their further
growth.
I now make the following possibly surprising assertion, which I hold
to be crucial:

Quantum mechanics, like a diver, can take us down to the level of the
very first actions of our conceptualization of reality. And starting
from there, it can induce an explicit understanding of certain fun-
damental features of the new scientific thinking.

The following remarks can give a first idea of the content of this as-
sertion. Our way of conceiving the “object”, which is what we separate from
the “rest” in order to enable us to definitely examine and reason about it,
marks our whole way of thinking as well as all our actions. Now, intuitively,
the word “object” is still quasi-unanimously felt to be essentially tied to
invariance, material, morphological, and functional, and thus to what could
be called an “intrinsic objectivity”, independent of observation, pre-existing
such as it is perceived. More or less implicitly, all of current language and
the entire classical logical and probabilistic thinking are founded on this pre-
supposition. But quantum mechanics opposes a direct, radical and definitive
xii

veto of this presupposition. If its cognitive strategy is fully decoded and con-
veniently generalized, the formalism of quantum mechanics acts like a strong
magnifying lens under which the static contour of the classical concept of
object dissolves into a complex process inextricably tied to human cognitive
actions, most usually reflex actions, but often also deliberate ones; and, in
any case, the result of this process is indelibly marked by relativities to all
the cognitive actions involved. In essence this conclusion has been known
well for a long time. But the specific way in which quantum mechanics
conveys this old conclusion is new, and it amounts potentially to an overt
seizure by physics of the basic metaphysical question of realism. Physics
thereby merges with philosophy in a basic, massive way, and it injects into
philosophy a stream of innovation that leads directly into epistemology :

Quantum mechanics has captured and represented—for the particu-


lar case of microstates and in an implicit, cryptic way, but for the
first time in the history of human thought and directly in mathemat-
ical terms—certain universal features belonging to the very first
stage of the processes by which man extracts chains of communicable
knowledge from the physical reality in which he is immersed and of
which he partakes.

This is what the epistemological universality of quantum mechanics


consists of. By no means does it consist, as is often asserted, of the fact
that any material system is made of microsystems—which is a physical cir-
cumstance, not an epistemological one. The feeling of essentiality conveyed
by the quantum mechanical formalism to those who can read it, does not
stem from this physical circumstance; it stems exclusively from the uni-
versal character of the peculiar cognitive situation dealt with in quantum
mechanics. And, while reflections of it are encrypted in the general fea-
tures of the formalism considered as a whole, this cognitive situation marks
also directly the specific formal features that are pointed toward by the ex-
pressions “quantum probabilities” and “quantum logic”. These simply are
not intelligible in terms of what is classically called probabilities and logic.
This manifests strikingly that the general epistemological consequences of
the quantum mechanical formalism, if elaborated, modify the structure of
our classical representations of probabilities and of logic, the two most basic
and worked out representations of domains of our everyday thinking and
acting. Indeed, when the universal representation of the very first stage
of our conceptualization processes, drawn by generalization from quantum
mechanics, is injected into classical probabilities and classical logic, they un-
dergo a sort of spectral decomposition; and this places into evidence that, far
down beneath language, probabilistic and logical conceptualization merge
INTRODUCTION xiii

into one unified probabilistic-logical structure. This circumstance entails


deep conceptual clarifications as well as corresponding formal modifications.
No other theory of a domain of reality, not even Einstein’s relativity, has ever
triggered an outflow of a comparable scope, so deep-set and so powerfully
innovating.
This, however, though variously felt and much discussed and analyzed
for more than 70 years, often with remarkable penetration, nevertheless is
still very far from being fully known and understood. The general episte-
mological implications of quantum mechanics are still cryptic, even for most
physicists and even for many who currently manipulate the formalism, often
in a masterly manner. A fortiori, quantum mechanics is very superficially
and feebly connected to the development of other new scientific approaches.
This is a huge lacuna. It hinders a free, rapid, and maximal development of
the revolution of the basic concept of object, implicitly started by quantum
mechanics, but the pressure of which manifests itself also in biology, sys-
temics, information theory, etc. Thus it also inhibits the perception and full
elaboration of the consequences of this revolution upon logic and probabili-
ties that guide our everyday thinking. Thereby it obstructs the now-possible
radical progress in our knowledge of our manner of producing knowledge.
Which furthermore delays a now-possible dramatic improvement of an ex-
plicit and deliberate domination of our epistemological behaviour, and thus
also of our actions.
One of the main aims of this book is to fill this lacuna.
This aim joins with a still larger one, which stems from the postulate
that any big theory of a domain of reality fixes in the concepts and the
structures defined by it, certain essential features of the epistemological
processes by which the human mind generates representations of what we
call reality. But, as happens in the special case of quantum mechanics, these
features tend always to remain more or less implicit in the descriptional
substance that has incorporated them, which entails that their universal
value remains unused. A fortiori, the different epistemological innovations
that accompany different scientific approaches, in general remain unreferred
to one another, which blocks the emergence of an integration.
So, for instance, the theory of information obviously involves a certain
epistemological universality. Any “transmission of knowledge”—even if it is
a natural, non-intentional process of just the acquisition of knowledge, or a
scientifically normed process of measurement, i.e., of deliberately organized
transmission of data from an object of study to the mind of an investiga-
tor, etc.—can be cast in the canonical mould of the theory of information,
according to which there always exists a source of “information” that issues
“messages”, a “channel” for the transmission of information which can alter
xiv

in various ways the messages sent through it, and a “receiver” that attempts
to restore the original message out of the received one. This remarkable gen-
erality entails a tendency to apply the informational representation (initially
conceived for the engineering of communication devices) to the most diverse
domains, in biology, in the theory of physical measurements, in linguistics,
and so on. It would therefore certainly be fruitful to explicate thoroughly
the general epistemological presuppositions of the information-theoretical
formalism and to confront them systematically with those involved in other
approaches. The theory of quantum mechanical measurements clearly of-
fers an opportunity for a particularly interesting confrontation. Indeed, this
theory distills the essence of fundamental quantum mechanics and quite es-
sentially addresses an informational problem. Nevertheless, the formalism
of the quantum mechanical measurement theory possesses certain formal fea-
tures that are essentially different from those of the informational for-
malism. It would by interesting to explore what facts, assumptions, and
methodological choices underlie this unexpected difference. While it might
produce a deeper understanding of the, so central, general concept of “infor-
mation”, such an investgation could perhaps furthermore lead to a reformu-
lation of the theory of information in terms of Hilbert mathematics,2 which
probably would be a formulation much deeper, more precise and general than
the present one. In turn, a re-expression in terms of Hilbert mathematics
of the theorems from information theory (especially the second theorem of
Shannon) could draw the famous question of hidden parameters into an
organized and mathematical framework; additionally it should foster impor-
tant clarifications concerning the concept of physical superposition as well
as throw further light on the concept of “object”.
Considerations of a similar nature could be advanced for several other
modern disciplines, in particular for the various computational approaches,
for molecular and genetic biology and, quite specially, the modern cogni-
tivistic approaches.
But the preceding considerations suffice already to convey the follow-
ing conclusion:
What is lacking in order to improve our knowledge and control of
the modes available for the generation and communication of knowl-
edge, thoroughly and rapidly and with precision and detail, is a
systematic research within the mutually isolated special languages
belonging to all the major modern scientific disciplines, of the epis-
temological essence inherent in every one of them, and a systematic
cross-referencing of the explicated results.
2
I do not write Hilbert vectors, because evidently a principle of superposition permitting
a pertinent use of vector spaces does not hold for any transmission of information
INTRODUCTION xv

Indeed, in its own sphere of representation, each approach traces a certain


specific direction of conceptualization. But what are the “angles” between
these directions? What are the contents of their “projections” on each other?
And what new metawhole can pertinently be constructed from such com-
parisons?
This conclusion and the questions that surround it lead us to formu-
late the following aim: from the most profound and best-performing modern
scientific disciplines, to induce an explicit and formalized method of concep-
tualization, basic enough to:

(a) encompass in a unifying and optimizing structure the main spe-


cific procedures for generating knowledge employed in all these disciplines;
(b) assign within this structure a definite location for each one of
these procedures;
(c) generate comparability among these specific procedures and
among their results.

This, I hold, is an important aim. A better understanding of it can


be gained from the following specifications:
From the start, what is desired is the construction of a method, not of
a neutral description of the processes of conceptualization such as they may
spontaneously emerge. In fact, a perfectly neutral description would not be
a possible goal, and, even if it were, it would be devoid of any definite and
immediate pragmatic interest.
As for the requirement of a “formalized” method, it can be explained
as follows: Any methodology involves its subjection to some system of aims.
A minimal finality that seems imperative when a method of conceptualiza-
tion is planned, is to offer general algorithms for excluding the emergence
of false problems and paradoxes, while insuring rapid progressions, with-
out hindering thereby a fully free exercise of the peculiar curiosities of the
conceptualizing mind. The foregoing, if at all achievable, can however be
realized only by an extraction of the method from the current language.
The most radical extraction would result from the definition of a “formal”
method where exclusively nonverbal symbols, well-formed sequences of such
symbols, and transformation laws involving all of these, are put to work. But
this is not the aim proposed here, because significance, semantics, is primor-
dially essential when one conceptualizes. So, instead of “formal”, we use the
term “formalized”, which implies that something to be formalized has been
formed before, independently (as, for instance, is the case for a mathema-
tized theory of a domain of physical reality, say, the Faraday-Ampère-Biot-
Savart-Laplace-etc. system of descriptions, which Maxwell then re-expressed
xvi

in mathematical terms).3 Accordingly, in our case the first stage should con-
sist of the explicit construction of a general system of posits, definitions, and
procedures, constituting a self-consistent network of routes for directed and
safe conceptualization, inaccessible to the innumerable and unpredictable
obstacles inherent in the tortuous paths of conceptualization which each
one of us hews for himself in accordance with his own ability and way of
thinking induced in his mind by the usual language. Of course, a system
of this kind has to be expressed in words. Nonetheless, as a system, it is
a self-consistent whole, already extracted from current language, already
endowed with a certain degree of imperviousness with respect to an uncon-
trolled inflow of harmonics of significance triggered by words depending on
the density of the structure the system has been endowed with.
The second stage, then, should consist of a formalization of the
methodological system constructed in the first stage (or in several formal-
3
From one contribution to this volume to the next, the reader will notice oscillations
between the terms “formalized” and “formal”. In this connection, in a recent letter, Hervé
Barreau wrote to me:
“ . . . As for the essence, we are in agreement, since for all of us, and especially for you
and me, it is quite obvious that the sort of epistemology we want to construct presupposes
that we conserve the (often very complex) semantic of the involved terms, upon which
we shall try to impose constraints of “form” in order to stabilize invariants of meaning
which in the usual language in general get lost. Initially, for me, “formal epistemology”
meant precisely this submission to formal constraints of a basic semantic which has to
be kept. What rejected me in the expression “formalized epistemology”, was that it
might be understood accordingly to the opposition between “formal logic” and “formalized
logic”. The formal logic, of which the classical example is Aristotle’s logic, conserves in
it a basic semantic which permits to produce counter-examples in order to exclude a
possibility that is allowed by the criteria of pure form: for instance, when he wants to
exclude certain syllogistic modes relative to some given “figure”, Aristotle gives proofs
by ecthesis, that is, by specification of a counter-exemple (this procedure is still current,
in particular, in modal logic). While on the contrary, formalized logic makes abstraction
of any content. This is not the case in Frege’s first presentation of his logic, but this
is the case in the axiomatization of his logic. This is equally the case in Wittgenstein’s
“semantic tables” where the only “semantic” notations kept (namely “true” and “false”)
finally are indifferent since the tautologies, the formal laws, are valid independently of
the truth-values of the involved statements. So the formalized logic concerns exclusively
statements and not propositions (statements asserted to be true or false). In a similar
way, for the formalists mathematics is a formalized knowledge that is independent of the
semantic content, not only a formal science. This is the distinction which I had learned in
the school books of logic and mathematics. But the explanation you gave last Thursday
assign an opposite significance to this opposition, and it raised no objections . . . .”
This quotation shows clearly that (a) throughout this volume it is admitted by all
the contributors that the semantic contents are an essential element in the researched
epistemology; (b) those who use the word “formalized” refer to the paradigmatic example
of a mathematical theory of a domain of physical facts, while those who make use of the
term “formal” refer implicitly to certain traditional expressions concerning logic (though
nowadays “formal logic” is considered to deal with purely syntactical systems).
INTRODUCTION xvii

izations), mathematical or not, the initial outline being kept present as a


nourishing soil. Thereby, without loss of nuances, the precision and effi-
ciency of the processes of general conceptualization achievable by use of the
method would become comparable to those which logic has attained for the
particular purpose of combining and transporting truth-values of proposi-
tions, or to those which a mathematical theory of a domain of physical reality
insures for the representation of physical phenomena, under constraints of
inter-subjective consensus and predictability.
A methodology of the kind specified above is what we call a
formalized epistemology.
By the nature of its aim, a formalized epistemology should emerge much
more general and, nevertheless, by far less abstract than the representations
built in metamathematics or in the logical theories of hierarchical languages.
The project sketched above should not be mistaken for a crossdisci-
plinary or a multidisciplinary project. The latter projects are designed to
offer to nonspecialists access to information, to results obtained inside spe-
cialized disciplines, as well as a certain understanding of these results; by
contrast a method of conceptualization should equip anyone with an instru-
ment for conceptualizing in whatever domain and direction he or she might
choose. Our planned method should furthermore not be assimilated either
with any approach belonging to the modern cognitive sciences, which try
to establish as neutrally as possible descriptions of how the human body-
and-mind function spontaneously when knowledge is generated; whereas a
method of conceptualization should establish what conceptual-operational
deliberate procedures have to be applied in order to represent and to achieve
processes of generation of knowledge optimized according to definite aims.
It seems however clear that a method of conceptualization of the
sort we have defined would share some features with the crossdisciplinary
or multidisciplinary approaches and with the cognitive sciences (as well as,
furthermore, with a theory of a domain of facts).

Now, is a formalized epistemology possible at all? The hopeful pur-


pose of this volume is to bring about agreement on a positive answer.

The volume is organized in three parts.


The first part offers various perspectives on the aim proposed in this
Introduction: its historical roots, its present conceptual environment, esti-
mations of its possible content and of its pragmatic value, the difficulties
entailed by it, and its a priori chances to succeed. These preliminaries seem
necessary in order to deepen the intuition for what is desired and to create
a background for the constructive attempts we will propose.
xviii

The second part contains three constructive approaches which form


the core of the present volume.
The third part features critical-constructive explorations concerning
the present stage of knowledge in several different domains of investigation
(philosophy of time, physics, logic, mathematics, computation, linguistics,
and complexity), each one more or less explicitly related to the concept
of a formalized epistemology. In this manner, around the constructive ap-
proaches from the second part, new ground is broken for future positive
developments.
The whole, I think, will offer a rather complete account of the syn-
thesizing dynamics conducted within the CeSEF.

Mioara Mugur-Schächter

IMPORTANT NOTE

For the reasons indicated in the above Introduction, please


read “formalized epistemology” instead of “formal epistemology”
wherever the latter term appears in Chapters 1, 3, 5, 6, and 8.
We much apologize to our readers for this unavoidable inconve-
nience.
Part One
Preliminary Explorations:
What? Why? How?
1
REMARKS ABOUT THE PROGRAM FOR A FORMALIZED
EPISTEMOLOGY∗

Francis Bailly

Laboratoire de Physique du Solide de Bellevue


CNRS, 1 Place Aristide Briand
92195 Meudon Cedex, France

The question of relationships between mathematical structures and language


analysis in epistemology is considered briefly in the framework of a program
for a formalized epistemology.

Key words: language, formalization, mathematical structures.

1. PRESENTATION

This short paper does not pretend to analyze either the full importance
or the stakes of a formalized epistemology such as the one proposed by
the CeSEF. We shall limit ourselves to pointing out a few tracks likely to
prove interesting to follow and to show the long-range aim and relevance of
such a project. My own position is determined, of course, both by personal
attitudes about general commitments (in philosophical, ethical, political do-
mains) and by a professional practice in research in physics, i.e., in a dis-
cipline where mathematics have proved to be both deeply explanatory and
fruitful in building new concepts and producing counter-intuitive notions. It
appears particularly that natural language and every day conceptualization
remain unable to account directly for physical features and properties, while
mathematical formalisms made them easily understood. On the basis of this
experience it is tempting to look at what, in epistemology and reflections
about scientific knowledge, can be defined and formalized in order to free
this specific domain from the over-determinations of natural language and
commonplace representations.


See “Important Note” on p. xviii.

3
4 F. BAILLY

2. WHY A FORMALIZED EPISTEMOLOGY?

Mathematical structures have their own developments, mutations and mu-


tual transformations. Strongly formalized physical theories remain “self-
sufficient” insofar as they have to describe and explain the material world.
Biological sciences are continuously forging their concepts and constructing
their objects and seem on the way to explain living organisms and their
intrinsic complexity. Even social sciences have developed to some extent
their own meta-languages about the problematic reality they are in charge
to treat. For its part, epistemology has succeeded in elaborating discur-
sive and conceptual methods that enable it to characterize and analyze the
specific ways of scientific knowledge. What needs, then, to be questioned
are the interest and the possibility for a “formalized epistemology” to ex-
ist. However, being inspired by the earlier movement of axiomatization in
mathematics and its consequences, we could retain at least three types of
considerations in order to justify pursuing such a program:
(i) It provides the possibility, through the requirement of some formaliza-
tion, to elicit many presuppositions and implicit postulates involved
in scientific theories as well as in the epistemological analyses linked
to them, thus helping to clarify the involved contents and procedures.
(ii) It makes it possible to bring into evidence the incompatibilities or
even contradictions contained in certain analyzes, which are difficult
to express through the pure discursive argumentation. It may thus be
used as a tool for criticism of interpretations and representations.
(iii) Thanks to the internal generativity of the formalisms themselves, it
might make possible the discovery of new ways of research, in the
same manner as the mathematical modelization of the phenomena do.
It may thus play a heuristic and fruitful role for analyzes. Beyond the
opening up of these possibilities, it is tempting to formalize epistemol-
ogy in a way that could lead to make more explicit and even to redefine
the role and the use of language in a theory of scientific knowledge.1
This point will be briefly considered in Sec. 3

3. NEITHER A COGNITIVE SCIENCE NOR A


REDUCTIONIST PROJECT

Indeed, before discussing this last point, we have to stress the fact that de-
spite the appearances and even if some similarities may be found, such a
1
In the same spirit but in a different manner as that which has been attempted by
some recent researches [1–3].
ABOUT THE PROGRAM FOR A FORMALIZED EPISTEMOLOGY 5

project is not equivalent to the development of a research program in cogni-


tive sciences, nor, conversely, to an attempt for a renewed logical reduction,
as did in their time the philosophers of the Wiener Kreis. Let us point out
the differences between these two perspectives.

(i) On the one hand, a formalized epistemology cannot simply be a part


of cognitive sciences to the extent that its investigation range does
not identify with constructing a “scientific object” as has to be the
case for cognition: as emphasized by G.-G. Granger [4], philosophi-
cal knowledge is a “knowledge without object,” and epistemology in
its philosophical version does not aim at constructing an object, but
rather at elucidating the processes of such a constitution in sciences.
To this assumption it could be objected that elaborating a formalism,
as epistemological as it would be, determines ipso facto some objec-
tivity as a correlate and a referent for this formalism if it appears to
be adequate. Answering this objection requires the notion of “formal
content” (as introduced by Granger [5] in his epistemological analysis
of the mathematical science), extending its relevance according to two
points of view: first, a formal epistemology might be considered as a
formalized epistemology of such formal contents and, second, it might
be considered as the research of the mutual articulations, in a given
scientific theory, of the formal contents this theory produces. Thus, if
a formalized epistemology leads to the rising of some “pseudo-object”,
the latter refers in fact to a mathematical universe of concept con-
struction dealing with the interpretation and mutual coupling of the
implied theoretical concepts. It follows that this “pseudo-object” re-
mains determined less by the formalization of the epistemology itself,
than by the scientific disciplines which have generated it.

(ii) On the other hand, it is known that the abstract logicism of the Wiener
Kreis, has to do with an empiricism as regards phenomena. It leads
to a quasi-ontological disjunction between two worlds: the one of logic
and the one of phenomena (considered as sets of data). Conversely,
a formalized epistemology would develop the aim to explore a unique
world: the one of the “scientific object” as such [6], through the anal-
ysis of its effective construction in the discipline where it is produced.
Formalizing this analysis would offer a double advantage: the first
one, already mentioned, is to detect through their traces the cognitive
operations making possible this process; the second one is to permit
the formation of a new meta-language regarding simultaneously both
these operations and the concepts they treat. With the hope to make
more evident the conditions of possibility for such a construction of
6 F. BAILLY

objectivity, reviving thus, but only to this extent, a transcendental


approach [8, 9].

4. FORMALISMS AND LANGUAGE

This point leads us to a quite general feature, which seems to be linked with
every formalization of knowledge. Indeed, as we stressed elsewhere [7], when
we have to deal with more or less formalized sciences, natural language ac-
quires two distinct functions, that formalism enables us to distinguish and
separate: a referring function and a referred function. In its referring func-
tion, language provides the means to express and establish the axiomatic of
the formalisms or the main theoretical principles underlying the discipline.
Somehow, it governs the objectifying activity. In its referred function, the
language uses more (technical) terms than (usual) words, more conceptual
relations than signification. It appears as submitted to the proper determi-
nations of the abstract structures it contributed earlier to construct. Until
new scientific theorization leads to use this referred state of language in order
to confer it a new referring function in view of new formalisms or new princi-
ples, more abstract or more general. And so on, from paradigm to paradigm,
from themata to themata, from epistemological cut to epistemological cut.
In this continually acting process, the formalism as such keeps the space
open as well as the splitting - which remain fundamental for constructing an
objectivity - between these two functions of the language, thus enabling the
mediation between them. More and more assured and in evolution, thanks
to the first, it modifies continually the second through its proper internal
dynamics, as is well shown through the intrinsic generativity of mathemat-
ics in modelizations. Meanwhile the formalism contributes in generating the
language through both the functions the latter has alternatively to fulfill
and between which the former assures a ruled communication.
In the usual practice of epistemological analysis, these two functions
are very feebly mutually individualized. Their relationships are deciphered
in the light of the conceptual analysis of scientific theories themselves and
the referring function is made use of in considering the new relations induced
by the formalisms in the referred function, while this last one is made use of
for putting into evidence the theoretical concepts involved in the formalisms.
In order to achieve this, epistemology calls for a philosophy of knowl-
edge, at the same time that it uses the disciplinary language with its own
concepts. Thus, to aim at a formalized epistemology amounts to aim at
reiterating the proper device of sciences, on the interpretative and compre-
hensive level, and at renewing its power of explanation. It also raises hopes
that such a reiteration would lead us to build some real formal hermeneu-
ABOUT THE PROGRAM FOR A FORMALIZED EPISTEMOLOGY 7

tics (in the same sense that mathematics can be regarded as such a formal
hermeneutics [10].)

5. CONCLUSION

Finally, and despite the fact that it is not the first time it has been tried
to build an approach deserving being called a formalized epistemology (cf.,
e.g., the above references to the Leibnizian “universal characteristic”, to the
logical research of the Wiener Kreis, to the set-theoretical approaches of
Stegmüller and Sneed), the program proposed by the CeSEF may appear as
a new possible progress in understanding the elaboration and the status of
scientific knowledge, as well as the role played by linguistic representations
and by language itself in their interpretation. Of course, it is out of question
to reduce the interpretative process to the benefit of some excessive “formali-
cism”, or to reduce epistemology itself to a purely cognitive approach. But,
according to the concern of Wittgenstein about philosophy, it could be seen
as a concern of freeing epistemology from some spontaneous determinations
induced by natural language, by taking into account scientific results: it
could play the role of some “therapy” (following Wittgenstein’s provocative
terminology) for a reflection about science and its cultural appropriation, as
well as a revival in the research for the deep cognitive invariant structures
underlying science.

REFERENCES

1. J. D. Sneed, “Philosophical problems in the empirical science of sci-


ence: A formal approach,” Erkenntnis 10, 115 (1976).
2. T. S. Kuhn, “Theory change as structure-change: Comments on the
Sneed formalism,” Erkenntnis 10, l79 (1976).
3. W. Stegmüller, “A combined approach to the dynamics of theories.
How to improve historical intepretations of theory change by applying
set theoretical structures,” in The Structure and Development of Sci-
ence, G. Radnitsky and G. Anderson, eds. (Reidel, Dordrecht, 1979).
4. G.-G. Granger, Pour la connaissance philosophique (Odile Jacob,
1988).
5. G.-G. Granger, Formes, operations, objets (Vrin, 1994).
6. F. Bailly, “Sur le statut contemporain du concept d’‘objet-
scientifique’,” Rev. Int. Syst. 9 (4), 385 (1995).
7. F. Bailly and J. Petitot, “Les mathématiques, de la diversité a
l’unification,” Encyclopedia Universalis, Symposium (1990), p. 700.
8 F. BAILLY

8. J. Petitot, “Logique transcendantale: Synthétique a priori et


herméneutique mathématique des objectivités,” Fundamenta Scien-
tiae 10, 1 (1990).
9. J. Petitot, “Idéalités mathématiques et Réalité objective. Approche
transcendantale,” in Hommage à Jean-Toussaint Desanti, G. Granel,
ed. (TER, 1991).
10. J.-M. Salanskis, L’herméneutique formelle (Editions du CNRS, 1999).
2
FORMALIZED EPISTEMOLOGY IN A PHILOSOPHICAL
PERSPECTIVE

Hervé Barreau

23 rue Goethe
67000 Strasbourg, France

The need for a formalized epistemology is recognized by all scholars who


think that the relativity of all sciences must not be referred to a social rela-
tivism. In XXth century, Husserl was the protagonist of such an epistemic
philosophy. But this philosophy was more successful in social and human
sciences than in natural sciences. In this latter domain, quantum mechanics
obeys the requirements of a Kantien perspective in a more precise sense that
was the case with Newtonian mechanics.

Key words: relativity, relativism, phenomenology, epistemology, quantum


mechanics.

1. INTRODUCTION

To my mind, what motivates thinkers of diverse schools and tendencies to


adhere to a project of a formalized epistemology is the fact that there is
no other plausible alternative, even if each one among us conceives such a
project in an original manner.
And why is there no other plausible alternative? Simply, I think,
because, should we not take heed, the task of epistemology would be more
and more taken up by authors who, although they have impeccable commu-
nication skills, possess a far less solid, not to say dubious grasp of scientific
knowledge. Over the last thirty years we have witnessed the publication
of works characterized by such loose accounts of the principles and results
of the theories of mathematical physics, that the very essence of these is
dissolved. The authors of such works distort and misrepresent the scientific
discourse, on the basis of the misguided conception that scientific theories

9
10 H. BARREAU

are a mere object of blind faith agreed upon amongst specialists, nothing
more. In their mind, the remarkable relativizing methods by which modern
physics succeeds to include into rigorous descriptional structures the subjec-
tive elements which, in a non-removable way, mark any piece of knowledge,
are identified with a wholly arbitrary ‘relativism’. The existence, also, of
constraints stemming from a source distinct from human minds, is entirely
overlooked.
On the other hand – besides a well known and widely-experienced
need to gain some well-constructed insight into the results of the mathe-
matical sciences – there emerges the new goal to furthermore extract useful
general epistemological methods from such an insight.
Under these circumstances, an epistemology that aims to incorpo-
rate the essence of the methods of modern-day scientific theories into pre-
cise, formalized, but nevertheless intelligible, general representations of the
processes of generation of knowledge seems to be the only genuinely accept-
able perspective. This essence cannot be left unexploited. And the slope
which, from the observational principles of relativity that found any inter-
subjective consensus concerning reproducible physical phenomena, leads to
mere relativism, must somehow be suppressed.
This slope is a very slippery one indeed. Nietzsche glided down to
its very bottom. He considered scientific adventure as only an avatar of the
search for power, which needs to be closely kept under supervision because it
is continually deluding itself and others into the belief that what is attained
was truth. Heidegger, in his own way, went down the same slope.
Husserl, on the other hand, continuing Kant, developed a view that
strongly opposed relativism.
In what follows, I would like to make some remarks concerning the
developments that withstood relativism and the limits of these. For, beyond
these limits lie the main conceptual rooms where a formalized epistemology
could now build useful new contributions.

2. THE HERMENEUTIC ROLE OF PHENOMENOLOGY

Edmund Husserl fully understood what systematic belittlement of science


relativism was to cause. The forms which it took in the thinking of his former
assistant, Martin Heidegger, yielded already a striking illustration. In The
Crisis of European Sciences and Transcendental Philosophy [1], which is his
philosophical testament, Husserl wrote (pp. 21-22):

“Today we are aware that the rationalism of the XVIII century –


its way of wanting to guarantee the solidity and the proper behavior
FORMALIZED EPISTEMOLOGY IN A PHILOSOPHICAL PERSPECTIVE 11

required for the European humanity – was a naı̈veté. Yet, together


with this naive rationalism, which is even contradictory if followed up
to its final consequences, are we obliged to abandon also the authentic
significance of rationalism? And what about a serious explanation of
such a naı̈veté and contradiction? Where is the rationality of this
irrationalism that is being vaunted and towards which some want to
compel us?”

In his manuscript, written after a lecture given in Prague in 1935 under a


slightly different title, Husserl shows that the naı̈veté and the contradictory
character of the rationalism of the period of Enlightenment consisted in its
objectivism (or its positivism, as one might prefer to say), i.e., in the fact
that it remained unaware of the subjective contributions to scientific objec-
tivities that these objectivities were presented as purely factual data to be
accepted as entirely ‘real’. As early as the XVIII century, trends have been
manifested toward relativizing scientific knowledge by taking into explicit
account also the unavoidable subjective features that mark it, and in this
respect Husserl pays homage to Kant who initiated the transcendental phi-
losophy, of which Husserl’s transcendental phenomenology can be regarded
as both a critical review and a development.
From the point of view of a formalized epistemology, which is ours,
Husserl’s critical assessment of Kant is particularly interesting, for it shows
well the trap where any type of formalism may become ensnared. Of course,
for Husserl, Kant had the merit of having proclaimed the insufficiency of the
rationalism of the period of Enlightenment, notwithstanding that he shared
the essential ideals of rationalism. However, according to Husserl, Kant did
so in a manner that remained dogmatic and even “mythical”. As it is well
known, Kant, while showing scant interest for common knowledge, imposed
very elaborate conditions upon scientific knowledge, but without really an-
alyzing the intuitive elements that he posited at the basis of any knowledge.
Whereas according to Husserl, transcendental life – endowed with an a pri-
ori character, and generative of ‘objects’ – is so rich and cumulative that it
possesses a depth into which it is possible to penetrate by an analytic effort.
There each subjective phenomenon appears to act, in turn, as form and as
substance for other such phenomena, by a process that has no end. It is
mostly in this sense, I think, that Husserl’s approach is so original. The
philosophical ideals nourished by Husserl, the aim that he wished to assign
to the field of philosophy, are particularly well expressed in the following
text (op. cit. pp. 127-128):

“As soon as, philosophizing with Kant, instead of starting from the
same point and continuing with him along his own path, we turn
12 H. BARREAU

back upon his assumptions to question them (Kantian thought also,


as any thought, makes use of certain assumptions considered to be
obvious and beyond questioning), so as soon as we become aware of
these as of mere ‘presuppositions’ and we consider them as endowed
with an own universal and theoretical interest, from that moment
on, to our greatest astonishment, there opens up for us an infinity of
ever-new phenomena belonging to a new dimension and that come
to light only if one realizes coherently the implications of significance
and validity of these assumed obviousness: an infinity, I say, for,
while we penetrate further and further, it appears that each one of
them, such as we find it by unfolding its meaning, and also such as it
is initially lived and given as just obviously being, carries already in
itself implications of meaning and validity of which the interpretation,
in its turn, leads to new phenomena, etc. These – wholly – are purely
subjective phenomena, but thereby one should not conceive them as
mere factualities, mere psycho-physical processes concerning sensual
data, they are on the contrary mind processes, which, by an essential
necessity, perform the function of constituting forms of meaning. But
this, they always realize starting from a definite ‘material’ of the mind
which always reveals itself anew, by an essential necessity, to be in
its turn some figure of the mind, but called to become a ‘material’
in its turn, that is, to function as the constituent of a further figure,
just like what lastly appeared as a figure, then became a material.
No objective science, no psychology, despite the desire to set
itself up as a universal science of the subjective, nor any philosophy
either, has ever posited as a theme this realm of the subjective, and
consequently, never really discovered it. Not even the Kantian phi-
losophy, even though it wanted to come back upon the conditions
of subjective possibility of the experimentable and knowable world.
This is a subjective realm totally closed upon itself, which, in its own
manner, exists, which – in an ever-inseparable manner – functions in
every experience, in every thought, in every life, and nevertheless has
never been perceived, never been grasped nor conceived of.”

So, according to Husserl himself, his specificity with respect to Kant lies
much in the structure and functioning assigned to individual subjectivities,
and in the consequences of these assignments upon knowledge.
The richness of the new field thus introduced by Husserl in the domain
of philosophy is attested by the impressive number of authors who have
cultivated it and have therefrom reaped a notable harvest. However it is
astonishing that all the various methods that have been tried out on this
FORMALIZED EPISTEMOLOGY IN A PHILOSOPHICAL PERSPECTIVE 13

field – which often intersect with one another – have failed to generate a
science, the science of which Husserl dreamt, that would lie at the bases of
all the sciences and act as a common foundation for them.
This situation, of course, might be provisional. On the basis of a
phenomenologically-induced psychology are now developed ‘cognitive sci-
ences’ for which the links with phenomenology have not been severed. And,
to my mind, there is no contradiction between a phenomenology that is
devoted to the study of the relationships between the strata of living con-
sciousness and, on the other hand, an analysis of these strata carried out
via the methods of ‘objective psychology’ and of neuro-biology. These last
methods suggest and finally will impose certain formalisms, just as in the
case of the other sciences. Phenomenology, then, will have to take care that
its own adductions, associated with the formal elements, be unified into a
coherent corpus where the contributions from each source shall stay inside
the limits of their own validity, as confirmed by specific experiments. But
such a role for phenomenology confronts a real difficulty: it possesses no
other weapon than that passed down by Husserl, namely a striving to de-
velop a direct introspective knowledge of the, anonymous, essential processes
from the human mind. These, in order to create concerning them some sort
of ‘normalized’ inter-subjective consensus, should somehow become commu-
nicable in a clear and standardized way. Phenomenology in this sense is
quite different from introspection in the sense of just consciousness of one’s
own psychological operations. This, however, does not entail that it became
a current reference in the principles of some established science. Piaget,
when he made use of the clinical method, practiced something that lay be-
tween introspection and what Husserl meant, but he refused to acknowledge
any link with phenomenology, as if any connection between his own genetic
psychology, and psychology in some other, new sense, were suspect.
I personally think that, in the long run, phenomenology will settle
into a role of the type indicated above, a hermeneutic role, of interpretation
of the principles and results stemming from the cognitive sciences. And in
this sort of role it will indeed some day accompany all the sciences of man
and of society, not only the cognitive sciences, since all these sciences can
acquire validity only under the condition of an intimate and clearly estab-
lished connection with human psychological functionings. This condition
should constantly be called to mind as a criterion of trial among the in-
cessant production of ‘objective’ studies which in fact lose their objectivity
in so far that the subjectivity from which they stem becomes a subjectiv-
ity of scientists where there subsists an only loose relation with the initial,
founding, individual subjectivity.
14 H. BARREAU

3. PHENOMENOLOGY AND THE NATURAL SCIENCES

For the moment, however, and whatever the future of phenomenology will
be, it is a fact that the Husserlian vision of phenomenology, in spite of
its undeniable impact upon humanistics, has practically no influence at all
upon the natural sciences. In particular, it does not explicitly mark math-
ematical physics (I put the purely logico-mathematical sciences in another
category). Strongly centered upon the individual subjectivities, Husserl’s
view concerning the relations of these with scientific constructs are weakly
worked out: in any case insufficiently for radically damming up the ten-
dencies toward relativism. This boundary of the domain where nowadays
Husserlian phenomenology does notably act, brings into evidence – beyond
it – a vast vacant ground on which could very usefully be constructed a def-
inite pattern of formal representations of the processes of conceptualization
mainly drawn from the methods practiced inside the modern mathematical
sciences.
Among the numerous difficulties that confront such a project, I would
like to mention a particular one that is totally different from that encoun-
tered in the field of human sciences. There, the difficulty was to make us
share an individual intuition. Here the difficulty resides in the fact that in-
tuitions are almost always misleading, as Bachelard liked to point out, and
that they constitute obstacles that have to be dissolved. The formalisms
of natural sciences are not simple extensions of the lived symbolism. They
involve purifications and abrupt innovations commanded from outside the
individual psychological sources.
Husserl is quite right when he notes that in Galileo-Newtonian me-
chanics a ‘blanket of ideas’ is thrown over a world that was pre-geometric and
already structured. Yet he fails to point out also that the reason why this
Galileo-Newtonian science managed to develop roots is that it had founded
itself upon postulates and conventions that were efficient but far from being
unquestionable, and which – as any postulate or convention – can be refuted
if reasons emerge for doing so. Poincaré brought this into evidence force-
fully during Husserl’s youth. But, while Einstein understood the message
fully, Husserl was not moved by Poincaré’s views. According to him, the
Galileo-Newtonian revolution consisted mainly in the mathematization of
the representation of nature, which, he moreover thought, was not in the
least inconsistent with the debate on determinism which in 1935 animated
the field of atomic physics (ibid. pp. 61-62):

“Galileo, who discovered – or, to be equitable toward his predecessors,


who completed the discovery of – physics, so of nature in the sense of
physics, is a genius, both a dis-covering and a re-covering one. He
FORMALIZED EPISTEMOLOGY IN A PHILOSOPHICAL PERSPECTIVE 15

discovers mathematical nature, the notion of method, he paves the


way for the infinite number of discoverers and discoveries in the realm
of physics. He discovers, as opposed to the universal causality of the
sensible world (regarded as an invariant form of the latter), which has
since been termed simply the law of causality, the “ a priori form
of the ‘real’ world” (idealized and mathematized), ‘the exact law of
legality,’ according to which every event of ‘nature’ (of that which is
idealized) has to obey to exact laws. All this is a dis-covery and a
re-covery which we have taken, up to this day, to be just the pure and
simple truth. For, in the field of principles, nothing has been changed
by the so-called philosophical revolution consisting of the criticism,
by the new atomic physics of ‘the classical law of causality’. Indeed,
in my opinion, despite all this novelty, all which is capital with regard
to principles, remains, namely: the mathematical character itself,
the nature expressible by formulae and to be interpreted only on the
basis of formulae.”

One cannot but agree with the assertion that classical physics mathematized
the representation of nature and that this was an essential step. Yet it seems
clear that the deep methodological significances involved in the mathemat-
ical formulae whereby our human knowledge of nature is expressed, should
not be overseen. The very core of scientific revolutions consists of what the
mathematical formulae express concerning the way in which ‘natural phe-
nomena’ are actively determined by the scientist. Since Husserl makes no
explicit remarks on this point, he seems to have underestimated the impor-
tance of the methodological content of the mathematical formulae from a
physical theory. These – apart from their obvious predictive-operational-
observational powers – also involve peculiar features of a procedural kind,
wherefrom their performative capacities stem. The elements from the phys-
ical world and those from our minds have to be combined in certain definite
ways in order to reach the type of efficiency that characterizes mathematical
sciences:

These ways impose r e l a t i v i t i e s and they exclude


r e l a t i v i s m.

Heidegger’s unfortunate formulation that “science does not think” is


strongly misleading. Mathematical scientists are the deepest and most cre-
ative thinkers of present times.
16 H. BARREAU

4. OPERATIONALISM AND QUANTUM MECHANICS

If indeed science were nothing more than a collection of operational recipes


for predicting, we would not even know how to make a creative, non au-
tomatic use of it. Nor would physicists be able to reconstruct a theory if
this somehow went lost. They would be comparable to a man who, in case
of need, is just able to follow step by step the indications from a cooking
book. Whereas a real cook, even deprived of his book of recipes, can always
prepare a wonderful meal because he understands cooking, and he even can
write a new book of recipes. So a formal representation of the processes of
conceptualization, drawn from the methods practiced inside the present-day
mathematical sciences, certainly cannot be reduced to operationalistic aims.
In what a sense, then, should it transgress such aims?
Concerning the positivistic requirement of a radical suppression of
any ‘ontological’ search in connection with the quantum mechanical formal-
ism, Mugur-Schächter [2] writes (pp. 179-180):

“But I hold that such a purging is at the same time impossible and
frustrating. Whether verbal reference is made to it, or not, ontological
back-up has been infused throughout the whole action of construction
of the quantum mechanical formalism, in particular by the fact and
the way of generating the studied states, and above all, by the choice,
for each observable X, of a name for its mathematical representation
O(X), and of a method M(X) for ‘measuring’ it. Indeed, why has been
chosen precisely that association name-mathematical-representation-
method-for-measuring, rather than another one? Each stage has been
based on non-declared models, that is why..... It is not the formalism
that imposes the choices of M(X) and A(X) (A(X): Apparatus for
measuring X (our specification)), it is the task of the physicist to
make these choices, outside the theory, confined to the more or less
explicit use of the intuitions and models he bears in his mind. And
all these ontological adductions have become incorporated into the
form and efficiency of the obtained algorithms. A certain ontological
content is there, dissolved and assimilated in the very algorithms,
inseparable and confirmed. So what is the point of juggling this
ontological content away from our final manner of speaking, thereby
breaking the bridges toward our own modes of mental action?”

To begin with, let us stress that in this context the term ‘ontological’ is used
in the sense of ‘methodological model’, which is quite different, if not even
opposed, to the classical philosophical sense of ‘a-description-of-things-such-
as-they-are-in-themselves’. Now:
FORMALIZED EPISTEMOLOGY IN A PHILOSOPHICAL PERSPECTIVE 17

It is precisely in the nature of the possible ‘ontological’ contents – in


the sense of a methodological model – associable with the quantum
mechanical formalism, that consists the break between classical and
quantum mechanics.

Husserl failed to notice the schism because he was too absorbed with re-
constructing ‘a world of life’ from behind a world of formulae in which he
only saw a ‘clothing’. Whereas, if correctly interpreted, the formulae of
quantum mechanics modify the ‘body’ itself, the object constructed by the
physicists for qualification. According to many physicists the quantum ob-
ject called ‘state of a microsystem’ is conceived of as involving an ‘essential’
probabilism. This – in so far that one has indeed to accept it – is the most
fundamental revolution that has ever taken place in the realm of physics
since Galileo. But even if the ‘essential’ character of the quantum mechani-
cal probabilities turned out to be avoidable, still

the way in which actuality and potentiality intertwine in quantum


mechanics – so the status of contextuality there involved – mark a
radical departure from classical physics.

Thus, even though quantum mechanics by no means abandons the mathe-


matical style of physics initiated by Galileo, its writings point toward con-
tents that are so different from those of classical physical theories, that the
terms of these, and a fortiori those of usual language, cannot be used any
more with respect to the quantum mechanical writings without thereby in-
troducing a flood of misunderstandings. So quantum mechanics brings in
truly new types of significance which, in certain respects that will have to
be specified closely, resist the scheme constructed by Kant (and by Husserl)
for the development of a physical theory.
Nevertheless – in a certain sense that is defined by Mugur-Schächter –
the Kantian category of causality can be maintained: it is possible to build a
certain ‘causalizing’ modelization, namely a ‘minimal’ one (that has nothing
in common, neither with the currently quoted deterministic character of
the Schrödinger evolution law, nor with the de Broglie-Bohm attempt, but
concerns exclusively the connection between operations, potentialities, and
processes of actualization of these, which characterizes quantum mechanics
(V. Fock has also expressed a very similar view)). This, she thinks, is of
fundamental importance, for the following reason (ibid., p. 184):

“Ontological models do not ‘exist’ outside ourselves, in space, like a


star, and they cannot be ‘discovered’. We forge them ourselves. And
18 H. BARREAU

our manner of forging, starting from our perceptions, an ontology


that shall please us, shall appease us, that shall place us in a posi-
tion of psycho-intellectual equilibrium – is causalizing. This feature
probably corresponds to certain optimalities of adaptation (possibly
maximisations of the rapidity and adequacy of our reactions to the
environment) in the absence of which the species would perhaps have
foundered. Which irrepressibly suggests some harmony with the un-
knowable reality-as-it-is-in-itself.”

By mentioning such a ‘harmony’ with ‘the unknowable reality-as-it-is-in-


itself’, Mugur-Schächter, as she admits herself, departs from both Kant and
Wittgenstein.

5. KANT AND WITTGENSTEIN COMPARED TO THE NEW


FORMALIZED EPISTEMOLOGY

Though, as I have said already, I fully agree that quantum mechanics does in-
deed impose a distance with respect to Kant, I cannot completely agree with
the reason for which Mugur-Schächter holds that such a distance emerges.
I think that it is important to stress this point, because it concerns a mis-
interpretation of Kant that is wide-spread among physicists. I quote again
(ibid. p. 184):

“According to Kant the properties of physical entities cannot be


known ‘such-as-they-are-in-themselves’. Nevertheless, he left us free
to conceive that our perceptions, so our knowledge, are produced by
these very properties, which the entities possess intrinsically in an
actual way. But, as we have seen, the formalism of quantum mechan-
ics interdicts even this concerning the state of a microsystem.1 Our
knowledge concerning the state of a microsystem, in so far that it is
possible to make it emerge and to ‘explain’ it, can only be conceived
in terms of properties that are actualized starting from the properties
of the studied state, but that are different from these.”

The interpretation of Kant’s view asserted in the quotation is abusive, I

1
“Intrinsically”, in Mugur-Schächter’s explanation of quantum mechanics, is used in
a sense quite different from ‘such-as-it-is-in-itself’ and furthermore, different also from
certain more curent but vague acceptation. It refers to the author’s concept of “intrinsic
metaconceptualization [3] (pp. 260-264, pp. 270-273), a second stage in the processes of
‘relativized conceptualization’ of which the first stage – always – is a ‘transferred descrip-
tion’.
FORMALIZED EPISTEMOLOGY IN A PHILOSOPHICAL PERSPECTIVE 19

think. For Kant there exist no physical entities, nor space and time, inde-
pendently of the view of the knowing subject. Kant did not leave us free
to conceive that the entities which we distinguish correspond to ‘physical
entities’. He rather forebade this. For him such entities would be noumena
accessible only by ‘intellectual intuition’, a kind of intuition of which, in
fact, we are radically refused possession. For Kant the only intuitions we
are capable of are in space and time, the forms of our sensitivity. Which
means that we experience intuitions of – exclusively – phenomena. Out of
these, by using categories projected into the pure intuitions of space and
time by way of what he calls schemata, we create ‘objects’ endowed with
properties and linked by causal connections. All this is our own work. What
does not come from us is solely the ‘matter’ of our sensitive intuitions. But
this ‘matter’ comes from the anonymous ‘thing-in-itself’, not from ‘physical
entities’.
So it can be said that quantum mechanics concurs with the Kantian
interdiction to posit our knowledge as being produced by real entities. But
there is disagreement with the way in which Kant permits the constitution
and mutual connection of physical objects within his ‘analogies of experi-
ence’ and ‘postulates of general empirical thinking’. For nothing, in Kantian
epistemology, authorizes the conception of a physical explanation in terms of
‘relative potentialities and actualizations’, such as quantum mechanics does
authorize. While an alternative ontology, founded on the acceptance of an
essential randomness, would be even further removed from Kantian episte-
mology: in this sense – contrary to what Husserl asserted – the criticism
of ‘the classical law of causality’, in so far that it is accepted, clearly is a
philosophical revolution induced by physics. In this respect Husserl appears
to have been insensitive to the novelties brought forth by the science of his
time, in consequence of which the mathematical style introduced by Galileo
is now endowed with a fundamentally new sort of bearing.
As to the distance with respect to Wittgenstein, it concerns the
famous remark that the limit of what can be said consists of what can
only be ‘shown’: Mugur-Schächter remarks that quantum mechanical ob-
jects – states of microsystems – cannot be shown, and nevertheless they
are (retroactively) said, ‘dicted’, and then even ‘predicted’. It seems that
the ‘grammar’ of quantum mechanics has not drawn upon it Wittgenstein’s
attention, nor that of his students, as far as I know.
So it can be concluded, I think, that quantum mechanics indeed re-
veals to us new forms of objectivity, totally unknown to classical science
and epistemology. Thereby it contains precious guiding lines for a formal-
ized epistemology. And I quite agree with Mugur-Schächter that the theory
of information should also be used as another most important guide. Given
20 H. BARREAU

its remarkable generality, its quintessential extract from the phenomena of


communication, its successful applications in genetics and molecular biol-
ogy, it would certainly be extremely enriching to carefully incorporate its
epistemic contents. I would like now to draw attention on a very important
point. In a certain sense each great philosopher, in his own time, could
be said to have perceived science by way of some epistemology, that is, by
insisting on the fact that science never is founded exclusively on experience,
that it also involves choices of types of explanation or ‘causalities’ (one might
have said ‘paradigms’, had not this word, under the pen of Thomas Kuhn,
taken a too complex and ambiguous significance). Now, this circumstance
might lead to a confusion with what is here called a ‘formalized epistemol-
ogy’. It is quite essential, I think, to hinder such a confusion from the start
on:

A ‘formalized epistemology’ in our sense is not an epistemology of


science in general or of some particular sciences. It is a formalization
of general epistemological methods, drawn from the most performing
nowadays sciences.

To close now, I ask: Would a formalized epistemology make the theories


converge? Personally, I have doubts, since, starting from its origins, the sci-
entific undertaking manifests much more a tendency toward divergence than
toward convergence, and it seems unreasonable to believe that an epistemol-
ogy drawn from it shall head in another direction. Yet if we managed to
explain the divergent tendencies of the multiple forms of our representations
of empirical reality, their resistance to converge, then the epistemological de-
velopment undertaken here would acquire not only scientific value but also
a philosophical one.

REFERENCES
1. Edmund Husserl, The Crisis of European Sciences and Transcendental
Philosophy, translated into French from German by Gérard Granel
(Editions Gallimard, 1976).
2. Mioara Mugur-Schächter, “Les leçons de la mécanique quantique (vers
une épistémologie formelle),” Le débat 94, March-April, 1997 (Editions
Gallimard).
3. Mioara Mugur-Schächter, “Spacetime quantum probabilities II: Rela-
tivized descriptions and Popperian propensities,” Found. Phys. 22 (2)
(1992).
3
FORMALIZED EPISTEMOLOGY, LOGIC, AND GRAMMAR∗

Michel Bitbol

Centre de Recherche en Epistemologie Appliquee du CNRS


1 rue Descartes
75005 Paris, France

The task of a formal epistemology is defined. It appears that a formal epis-


temology must be a generalization of “logic” in the sense of Wittgenstein’s
Tractatus. The generalization is required because, whereas logic presup-
poses a strict relation between activity and language, this relation may
be broken in some domains of experimental enquiry (e.g., in microscopic
physics). However, a formal epistemology should also retain a major feature
of Wittgenstein’s “logic”: It must not be a discourse about scientific knowl-
edge, but rather a way of making manifest the structures usually implicit
in knowledge-gaining activity. This strategy is applied to the formalism of
quantum mechanics.

Key words: quantum mechanics, logical structure, Wittgenstein, philosophy.

1. INTRODUCTION

What makes possible the background against which is set our knowledge
of the order of nature? “There,” said Kant, “solutions and answers are
brought to a halt; because we must always go back to (this background) for
all answers and all thought of objects” [1].
At least this setting of limits, typical of transcendental philosophy,
points towards that which a formal epistemology cannot be. It cannot be
the formalism of an objectified theory of knowledge which would take the
subject-object relation as a second-order natural object, and would then
leave unquestioned the grounding of normative presuppositions on which all
science, including epistemology itself, depends. Further, a formal episte-

See “Important Note” on p. xvii.

21
22 M. BITBOL

mology cannot conform to the definition “of a clearly too foolish ambition”
which H. Putnam depicts as “[...] a superb theory of the normative grasped
in its own terms” [2] ; a sort of redoubling of the realm of norms of thought,
by which the theory would try to explain itself in objectifying the system of
its own principles, without being able, except by an infinite regression, to
question itself in return on its use of those very principles. Of course, these
remarks are not to deny the current attempts at naturalising epistemology
any interest. They are only aimed at pointing out that naturalisation of
epistemology can only be a process with no foreseeable completion; and that
at the provisional end of each step of this process there is a set of non-explicit
norms of investigation which we can but call the “pragmatico-transcendental
background” of the current state of research.
It is of no apparent advantage either to cast formal epistemology
in the role of a mathematicised or logicised variant of epistemology in the
modest sense traditionally intended in France: that of a multiplicity of crit-
ical analyses of the premisses and results of particular sciences. Because in
epistemology, as in the sciences, formalisation consists in the abstraction of
particular contents in order to reach universal rules. A formal epistemology
must therefore be of value to any science, even if it is especially profitable (as
we shall see below) when elaborating on the knowledge acquired by certain
methodologically advanced sciences.
Having discarded some tentative definitions of formal epistemology
it remains to give it a plausible positive characterisation. To that end a
comparative rather than a directly constructive strategy will be used. A
parallel will be established with the case of logic; the remarkable isomor-
phisms between logic and what we would expect of a formal epistemology
will be underlined; then, at the end of the discussion, the bringing into con-
sideration of some major differences between the two disciplines will allow
the formulation of the specific project of formal epistemology. It will thus
appear that formal epistemology can be understood as a generalisation of
logic; a generalisation of considerable range because it principally consists
in recognising the expansion of the form of the sciences beyond the closed
domain delimited by the Logos, taken in its narrow sense of explaining by
means of discourse consisting in predicative judgments.

2. LOGIC AND FORMAL EPISTEMOLOGY

Let us start from the dualistic prejudice of the theory of knowledge; be-
cause it is by way of criticising it that we will most quickly arrive at the
point of neutral equilibrium where both logic and formal epistemology stand.
Knowledge, according to Piaget [3] , consists in a certain relation between a
FORMALIZED EPISTEMOLOGY, LOGIC, AND GRAMMAR 23

subject and an object. It manifests itself by way of judgment (which consists


of ascribing a predicate to an object) or more generally by way of thought
shaped by the structure of judgment. Each science can be said to attain
knowledge, in the local sphere of objectivity to which it is assigned, if it ex-
presses itself via a certain network of judgments whose interdependence and
coherence fence off the temptation to systematically resort to ad hoc expla-
nations. But in that case, logic, which we traditionally present as a general
doctrine of judgment and of relations between judgments, is at once “sci-
ence’s doctrine”; logic, as Husserl points out, “[...] aims at bringing to light
the essential forms of knowledge [...] as well as the essential presuppositions
with which its forms are linked” [4] . In that, at least, the programme of
logic covers exactly the programme of a formal epistemology. In the dualistic
framework provisionally adopted, however, logic (and the formal epistemol-
ogy which matches it) takes on a sort of constitutive ambivalence. Logic and
formal epistemology are what Husserl calls “double-sided” disciplines [5] ;
disciplines having at once a subjective and an objective side. They have a
subjective side because they seek to extract the necessary states and regula-
tive principles of a “rational subjectivity in general” [6] . And they have an
objective side for two reasons. Firstly, because they engender ideal objec-
tive formations, as the product of their work of abstraction and deductive
generation; and secondly because, although they situate themselves below
the level of the concrete determination of objects and of classes of objects
of the particular sciences, they relate to the form of judgments, of which
the prime function is to characterise objects. It is this last thought which
led Husserl to characterise formal logic as “a priori formal doctrine of the
object” [7].
The two orientations of logic (subjective and objective) were given
privileged roles in turn by the actors of the history of philosophy. But this
process only led one to show the inadequacy of each single orientation as a
paradoxical result of the attempts to assure it the exclusivity. Let us con-
sider a first example. In the framework of the critical philosophy of Kant the
distinction between the reflexive and objective orientations of a discipline
does not rest on an exterior account of the face-to-face subject-object, but
rather on an internal analysis of the conditions for the possibility of experi-
ence. The “fundamental proposition” [8] of the critical philosophy effectively
announces that “The conditions of the possibility of experience in general are
at the same time conditions of the possibility of objects of experience” [9]; so
there can be no question of a confrontation between a pre-constituted sub-
ject and object, but rather a co-constitution of experience and its objects.
That being allowed, we notice that Kant’s internal analysis gives him two
motives for insisting on the reflexive orientation of logic. Firstly, logic situ-
24 M. BITBOL

ates itself entirely on the side of the formal aspect of our knowledge, without
any reference to its material and “objective” aspect [10] . Thus, according
to Kant, logic somehow situates on the “subjective” side of knowledge. Sec-
ondly, logic is freed from any link with the form of the sensible intuition, by
which “we can perceive objects a priori” [11] ; it proceeds without consid-
eration of perceived objects, turns on the understanding alone, and consists
in announcing the formal laws necessary to the thought of any object [12].
It is certainly not subjective according to the psychological conception of
subjectivity, because it is not content to describe in empirical terms the
intellectual mode of functioning of particular subjects; but we can call it
“reflexive” in so far as it is linked to the principles which order the thought
of the subject in general. Logic relates not to particular subjective facts but
to the norms which bear on the intelligent activity of any subject.
In that way, Kant does not just oppose logic to psychology, but also
to transcendental philosophy in its entirety. Because if transcendental phi-
losophy also deals, like logic, with the a priori formal conditions of thought
of objects, it does it via a very broad perspective in order to elucidate the
connection between knowledge and the faculty of knowing in general [13]; on
the contrary, logic is supposed to content itself with furnishing a “canon”
of agreement of one of the constitutive elements of the faculty of know-
ing (understanding) with itself. “In logic” writes Kant, “the question is
uniquely: how does the understanding know itself ?” [14]. In the evolution
of the Kantian project after Kant, the joining of form with the subjective
side of knowledge, the stratification of the sensible and intellectual and the
limitation of logic to a task of self-validation of the operations of the under-
standing, have been regarded as the weakest and least indispensable aspects
of the project’s completion. With the impetus of Cassirer and the many
protagonists of the “linguistic turn”, the integrated forms of symbolic ex-
pression have replaced the hierarchical forms of the faculty of knowing in
the role of preconditions of objectivity. Since then, as G. G. Granger points
out, the opposition between logic and transcendental philosophy has had
no raison d’être : “logic can [...] appear to be the most elementary and
the most radical aspect of the transcendental” [15]. “Logic is transcenden-
tal” [16], writes the early Wittgenstein, and, in the intention which it shares
with logic, formal epistemology is too.
In opposition to this process of abstraction and identification of the
Kantian a priori forms with the symbolic, another current of thought has
tended to put them into relation with the concrete turning points of the
phylogenesis and ontogenesis of the human subject. A psychogenetic rein-
terpretation of the Kantian hierarchy of the constituents of the faculty of
knowing has been proposed by Piaget, for example. According to Piaget,
FORMALIZED EPISTEMOLOGY, LOGIC, AND GRAMMAR 25

the underlying structures of natural thought issue from the stepwise coor-
dination of the diverse operational activities of the subject in the world.
But contrary to physics, which partially takes its information from the ma-
nipulated objects, by way of perceptual or experimental phenomena which
are supposed to open access to objects, logic proceeds from the exclusive
cordination of the actions which impinge on these objects and transform
them [17]. To the Kantian duplex of sensibility and understanding there
corresponds here a duplex of sensible receptivity and structured motor ac-
tivity of which only the second term concerns logic. According to Piaget,
“That which is axiomatised by formal logic is certainly an activity of the
subject” [18]. More precisely, it is a systematic activity of the subject whose
psychogenetic evolution has passed an essential stage: the conquest of the
reversibility of operations, which allows their formalised outcome to consti-
tute an ensemble of timeless and necessary connections [19].
But doesn’t exclusive concern with an ideal and isolated subject in
evolution keep us within a framework too narrow to yield reasons for the
emergence of the norms of thought? Does it not mask other genetic com-
ponents which are indispensable to the shaping of a logic? These additional
components are not denied by Piaget, but they do not constitute, in his
work, the material of systematic development. They concern just as much
the social interaction between subjects as that which is presupposed by them
regarding objects. On the one hand, although it is true that the construction
of logic is in the first place, according to Piaget, the work of a subject in
activity, its operational structures require “the collective contribution” of
other communicating subjects in order to be “reinforced and multiplied” .
“Reinforcement” ends in the stabilisation of norms by means of the symbols
used to communicate them. And “multiplication” could well refer to the
construction of non-classical logics which, not contenting themselves with
stating the normed forms of the effective operational activity of subjects,
formalise many possible operational activities by exploiting the supplemen-
tary free space which is offered by symbolism. The orientation towards a
theory of communicational intersubjectivity, favoured by contemporary Ger-
man philosophers such as Apel and Habermas [21], is thus able to complete
and enrich the focus on this work of “inquiry” of a generic subject which, be-
fore the work of Piaget, already formed the principal theme of many currents
of American pragmatism of the beginning of the 20th century [22].
On the other hand, the throwing back of the Piagetian problematic of
the normed activity onto the subject(s), its liberation in regard to the object,
calls at least to be nuanced. Activity is certainly, in the first instance, that of
the subject, but what about its regulatory forms which alone concern logic?
Piaget admits that it is “[...] impossible to know in advance if (these forms)
26 M. BITBOL

belong to the subject, to the object, to both, or solely to their relation” [23].
It must not be forgotten that activity consists in operations-of-the-subject-
on-objects. Even if it is indispensable to remove from the description of
this activity any mention of the particular features of the handled objects,
it must be recognised that the activity and its formal sediment rest on two
suppositions which bear on the objects in general: the supposition of the
permanence of objects and that of a minimal degree of stability in their
properties.
Let us dwell a moment on these two elementary suppositions of op-
erational activity, because they will have particular importance in the rest
of this paper. What must be noted about them from now on is that they
correspond term for term with those which the very use of the proposition
implies (by way, respectively, of the two fundamental acts of reference and
predication). For that reason, the formal kernel of the coördination of the
operations of the subject in the world corresponds closely to the formal ker-
nel of language. And so we understand that the axiomatisation of the motor
activities of the subject, on which Piaget focuses, converges at once towards
an axiomatisation of discursive activities to constitute that which can be
called a Logic. Nevertheless, one must bear in mind that the circumstances
which allow this remarkable convergence between the norm of the activity
and the norm of linguistic activity, are very peculiar. They are linked to
everyday life and speech. This urges us to introduce a reservation from
now on: nothing guarantees the durability of the relation obtained between
the domain of activity and the domain of discourse when we pass from a
gestural activity exercised at the heart of the everyday environment to an
experimental activity aimed at exploring its limits.

3. LOGIC, GRAMMAR, AND FORMAL EPISTEMOLOGY

On the other side of the dualistic demarcation, to wit according to the


philosophers who have privileged the objective side of logic, symmetrical
difficulties have provoked a swing of the pendulum back towards the same
point of equilibrium.
For the Wittgenstein of the Tractatus, the status of logic is that of
a representational framework. Logic, he writes, is a picture which reflects
the world [24], its propositions represent the scaffolding of the world [25];
“logical pictures can depict the world [26]” . Accordingly, logic merges
with the limit of the world. Indeed, the form of representation cannot for
its part be represented in the logical picture; it can only be shown by it
[27]. This remark, directed against the possibility of an authentic meta-
representation, justifies in its turn the crucial distinction which Wittgenstein
FORMALIZED EPISTEMOLOGY, LOGIC, AND GRAMMAR 27

makes between concepts and formal concepts [28]. We can say (with the
help of a proposition) that something falls under a concept, but we can only
show that something falls under a formal concept. “Object” and “property”
are such formal concepts. That something is subsumed under them is not
claimed but shows itself by way of the type of sign which is used to denote
it, or by means of the position of that sign in the propositional network.
The formal concepts of object and property are, so to speak, structurally
presupposed by language.
But as it is well known, that statics of mimesis typical of the Tractatus
is exactly one of the principal targets of the Wittgenstein of the Philosophical
Investigations [29]. The meaning of a proposition no longer establishes itself
in a projective relation to the world, but in a lateral relation to its use in a
“language game” or to its being put to work as a moment of a “form of life”.
The dynamic of this use appear to be constrained to a body of rules which
we call grammar in a wide sense; but it must not be believed, Wittgenstein
insists, that those who speak and act in conformity with those rules are
guided by them. Grammar is only the formalised residue of the practice
of language-games. In order to idenfy this formalised residue, one can rely
on the so-called “hinge” propositions of language. These propositions are
“[...] devoid of content because they do not admit of a negation endowed
with meaning” [30]. In other terms, they are devoid of content because they
constitute the minimal basis of tacitly accepted affirmations in relation to
which the affirmation or negation of all other propositions makes sense.
At this stage, if we would situate the thesis of the Philosophical In-
vestigations in terms of the dualism in the theory of knowledge, we would
need to ask ourselves some second-order questions: from what emerges the
symbolic practice with which that thesis deals? Is it imposed by the sub-
ject or the world? The later Wittgenstein turned at length around these
questions, but it was more to denounce their double lack of relevance than
to answer them. Even though practices are proper to the subject, they do
not reduce to series of arbitrary gesticulations and vocalisations. And even
though the grammar of practice is constrained by some “reality”, it does
not constitute a copy of this reality [31]. To paraphrase a remark of J. Bou-
veresse’s [32] concerning arithmetic, we should say thatthe connection which
exists between grammar and reality is something which can only be shown
in the application of grammar, and so it must not be described in terms
of correspondence with the facts accessible from a point of view exterior to
the practice of language-game. Just like logic in the Tractatus, or formal
epistemology according to the sketch which we have traced, grammar in the
Investigations is thus transcendental. It simply is so in a quite particular
way; not in the rigid style of Kantian a priori or the pictorial skeleton of
28 M. BITBOL

the Tractatus, but in the mobile manner of the functional a priori of Dewey,
qualified as quasi- a priori by Putnam [33].
Furthermore, grammar retains a feature which we have so far reck-
oned as characteristic of logic : the merging of presuppositions of discourse
and of action. “I act with complete certainty” [34] in accordance with the
norms which underpin the forms of life in which I participate; I speak in
complete confidence within the framework of rules of language of which I
make use; and I hold as unquestionable that background of propositions
“against which I distinguish between true and false” [35]. Forms of life,
background of “beliefs”, and rules of use of language, constitute for the
later Wittgenstein a coherent and undivided communal basis.
It is now possible to clarify the project of a formal epistemology by
means of a detailed play of similarities and differences between it and logic
and “grammar” in the sense of the later Wittgenstein.
To begin with, we have said that logic, “grammar” and formal epis-
temology are all transcendental. That is to say, they tend to reveal the
formal conditions of possibility for a state of knowledge (or a practical and
verbal orientation in the world). From that standpoint, they constitute
second-order disciplines, as against the first-order disciplines which are the
sciences. But they are not, for all that, meta-sciences or meta-theories rel-
ative to the theories of particular domains of objects. They take as their
object of investigation neither the sciences nor knowledge-gaining activity
as a whole. They content themselves with codifying a symbolising proce-
dure, and thus are able to make manifest the structures usually implicit
in knowledge-gaining activity. They say nothing; they show. They do not
represent; they present. They are typical examples of what G. G. Granger
very appropriately calls non-meta-theoretic meta-disciplines [36]; examples
of disciplines which in coming after a discipline do not establish between
them and the discipline a distancing relation as between a science and its
objects.
Further, we have underlined a considerable difference between logic
in the sense of the Tractatus and “grammar” in the sense of the Investi-
gations. The former has the rigid and hierarchical character of a structure
which presents itself as grounded in the giving of a world of which it ex-
hibits the “scaffolding”. “Grammar” has the mobility of a system of rules
following the lines of force of an interlacing of operational and linguistic
practices which is certainly constrained in some way by the real “other”,
but which has the elasticity to modify both the mode of expression of this
constraint and its mode of response. If we want formal epistemology to be
able to meet the challenge of scientific revolutions with a sure hand it must
resemble the “grammar” of the later Wittgenstein in its elasticity. It must,
FORMALIZED EPISTEMOLOGY, LOGIC, AND GRAMMAR 29

like “grammar”, belong to the class of evolutionary and non-foundational


meta-disciplines.
Finally, there is a common feature of logic and “grammar” which we
are inclined not to ascribe to formal epistemology; it is the presupposed
certainty of a concordance between the form of discursive practice and the
form of operational practices. This additional degree of freedom should al-
low formal epistemology to take care of a situation like that which confronts
quantum mechanics, in which there is no obvious agreement between the for-
mal coördination of operational activities and the structure of language. It
justifies in every way our calling formal epistemology an evolutionary meta-
discipline leaving in suspense the linguistico-operational concordance. And
it also justifies the expression “formal epistemology” when one attempts to
display the formal structure of a physical theory, as opposed to Y. Gauthier’s
expression “internal logic”. We can sum up these thoughts in the following
table:

Meta-disciplines Logic “Grammar” Formal Epistemology


evolutionary “Grammar” Formal Epistemology
lacking linguistico-operational Formal Epistemology
concordance

4. TO ACT BEFORE PREDICATING

The idea of a meta-discipline leaving in suspense the linguistico-operational


concordance is not entirely new. It is brought out very well, albeit in negative
relief, in a critique which Husserl addresses at formal logic. Formal logic,
Husserl explains, is of value for “[...] a real world thought of as already given
beforehand”. In traditional logic, the predicative structure of judgment,
together with the presupposition of the permanent existence of that of which
something is predicated, “[...] was self-evident and was never examined”
[37]. This constitutive pre-judged is equally brought to light by M. Mugur-
Schächter when she emphasises that language, logic and the classical theory
of probability rest on the common postulate of an “[...] intrinsic ‘objectivity’
which would prexist all acts of observation and conceptualisation” [38].
In contrast, Husserl proposes to go below the categorical structures
of language, below the form of judgment and below the formal concepts
of object, property or relation. In Formal and Transcendental Logic, and
more systematically in Experience and Judgment, he undertakes to put
“[...] in question their innate production and their springing up in the lower
stage of knowledge” [39]. Husserl calls this lower stage of knowledge “ante-
predicative experience”; and he shows page after page, with all the refine-
30 M. BITBOL

ment of his specific analyses, how from this can emerge the characteristic
moments of predicative judgment. The emergence of the substratum of pred-
ication and that of the predicate arises respectively by way of two modes of
ante-predicative experience: the “identificatory aiming” and the “explana-
tory experience” [40]. The identificatory aiming synthetically unites the
multiplicity of perspectives, of profiles or of aspects presented by perception
in an open experience of same and constitutes the precondition of the act
of reference to an identified object. As for the explanatory experience, with
its anticipatory tension, with its way of projecting interest towards the as-
pects which we expect to find if we modify our point of view on one and the
same object, it is at the basis of predication. An anticipation attested and
confirmed by the reproduction of a phenomenon when a certain perspectival
situation is reiterated can, in effect, translate into a predicate assigned to
the object aimed at.
But what would be the result if the phenomenological circumstances
of this twin stabilisation, of predicate as well as substratum of predication,
were not realised? What would happen in circumstances of the total disorder
of “ante- predicative experience”? Nothing less than this would result: the
disappearance of the conditions of an objectifying discourse making use of
predicative judgment; and consequently the equivocation of the means of
saying what this would amount to.
This aporia of the inexpressible can nevertheless be defused on two
conditions (which are not mutually exclusive):

(1) If the disorder of the experience is only partial and not total; because
in that case the failure of the anticipations associated with aiming at
an object could simply point towards the opportunity to change the
type of object or quite profoundly modify the mode of aiming towards
it.
(2) If, in response to the disorder, we can limit ourselves to a calling into
question of the lower levels of a logic which be less universal than that
which Husserl proposes.

The restricted calling into question which we propose is certainly super-


ficial when compared to the programme of genesis and foundation which
the creator of phenomenology formulated; but it is sufficient to deal with
the difficulties of contemporary physics. Instead of opposing, like Husserl,
the pre-logical circumscription of ante-predicative experience to the domain
where (predicative) logic is to be established, we would oppose the combined
domain of everyday life and instrumental operations, where the validity of
logic remains unquestioned, to the domain of the putative objects of exper-
imental investigation, in which the relevance of logical structures remains
FORMALIZED EPISTEMOLOGY, LOGIC, AND GRAMMAR 31

an open question [41]. If we proceed in this way, the loss of the conditions
for an objective mode of expression using predicative judgments within the
particular domain aimed at by experimental investigation does not have as
a consequence a global slide into the inexpressible, but only the restriction
of the sphere of relevance of the forms of discourse to the description of
instrumental operations.

5. QUANTUM MECHANICS AND FORMAL


EPISTEMOLOGY

In quantum mechanics we are exactly at this point. On the one hand the tra-
ditional forms of discourse using predicative judgments remain valid in the
domain of instrumental operations; better, they must remain so in as much
as they are the bearers of the preconditions of an intersubjectively shared
experimental knowledge [42]. But on the other hand, the expectations which
are induced by the aiming at a traditional type of object (corpuscular bearers
of properties) beyond the experimental apparatus, are generally confounded.
The expectation of being able to re-identify a corpuscular object founders
on the impossibility of experimentally following its trajectory continuously,
and on the indirect consequences of this impossibility (i.e., combinatorial
and statistical consequences). The expectation of seeing a phenomenon
reproduce itself is for its part systematically confounded in certain well-
documented cases: a value of a variable is not reproducible if, between two
occurrences of its measurement, we insert a measurement of a variable called
“incompatible” or “conjugate” (e.g. position and momentum).
So none of the phenomenological criteria for reference to a corpuscu-
lar type of object, and for the predication of determinations to that type of
object, are satisfied in the experimental domain of microscopic physics. We
are left in the presence of something which prima facie resembles an isolated
flux of singular experimental results, indissolubly dependent on the exper-
imental conditions which have given rise to them. In effect, these results
do not have a sufficient degree of invariance with changing experimental se-
quences for us to be able to detach them from the instrumental context of
their occurrence and to treat them as if they were evidence of a determi-
nation which an object would possess. In short, the events of microscopic
physics are essentially contextual, or again, as M. Mugur-Schächter says,
they arise from a “descriptional relativity”.
What is to be done in facing this critical situation, in which the cor-
roborated theoretical anticipations of the results of operational activity do
not satisfy the presuppositions of discourse reflected by traditional logic?
The first strategy, urgent and clarificatory, consists in showing, in manifest-
32 M. BITBOL

ing, the coördinated structure of these anticipations as it is extracted by the


formalism of quantum mechanics in a rigorous but not very explicit way. It
consists in capturing the meta-contextual structure which P. Heelan spoke
of [43], or the algebra of observation which S. Watanabe developed [44],
or the ordered system of relativising glances in the sense of M. Mugur-
Schächter [45]. To summarize, the strategy amounts to extensively utilising
the resources of a meta-discipline freed of the constraint of an isomorphism
between language and operations. A meta-discipline which corresponds ex-
actly to the definition which we have given to formal epistemology.
As a second strategy, we could always ask ourselves if it is not possible
to go back to the golden age of the linguistico-operational concordance in
changing logic (“quantum” logics), in choosing a new partitioning of the
world into objects having nothing in common with the material bodies which
bear localised properties (e.g., the referents of state vectors, as according to
Schrödinger [46]), or in assuming (as in hidden variable theories) that the
properties of corpuscular objects are instantly influenced by the instrumental
or environmental conditions of their manifestation [47].
But none of these second-level endeavours will be able to ignore the
lesson to be drawn from the first-level analysis brought to fruition by formal
epistemology. Quantum logicians face considerable difficulties in defining
what they mean by “property of an object” without conceding too much to
contextuality; the new ways of partitioning the world (i.e. the “new ontolo-
gies”) remain reliant on a level of discourse where a tacit “natural ontology”
operates; and hidden variable theories must have recourse to the artifice
consisting in substituting “contextualism” for contextuality: that is to say,
invoking a holistic influence of experimental circumstances on the underly-
ing intrinsic processes, rather than drawing directly the consequences of the
co-definition of the phenomenon and the conditions of its manifestation.

6. EPILOGUE

In the manner of the Euclidean geometers of Michel Serres, the physicist of


the classical epoch “[...] hopped on the moving train, at a moment when ev-
erything was already worked out, when the concepts were a thousand times
over- determined” [48]. Then, without clearly understanding what he was
doing or why he was doing it the physicist of the 20th century adopted the
path of a radical rexamination of the previously unquestioned articulation
between the operational and discursive domains. To that extent he puts
himself in the predicament of the modern mathematician who, in a para-
doxical development, “ [...] steers himself towards his unforseeable horizon
and his starting point” [49]; a mathematician who, to put it differently, ap-
FORMALIZED EPISTEMOLOGY, LOGIC, AND GRAMMAR 33

proaches more and more the performative origins of his science whereas he
thinks he gets closer and closer to his object. The meta-disciplinary analysis
of his science in the framework of a formal epistemology is apt to make the
contemporary physicist conscious of this reflective task which he has under-
taken in the wake of the mathematician, so clearly that nothing can ever
force him to fall back into forgetfulness.

REFERENCES

1. I. Kant, Prolegomena zu einer jeden kunftigen Metaphysik, die als Wis-


senschaft wird auftreten können, 1783, §36. A standard English transla-
tion can be found in: I. Kant Prolegomena to Any Future Metaphysics
that Will be Able to Present Itself as Science (Manchester Univer-
sity Press, 1971). As J. Petitot pointed out, Kant however modified
this very strict transcendental standpoint in his Opus Postumum: J.
Petitot, La philosophie transcendantale et le problème de l’objectivité
(Osiris, 1991).
2. H. Putnam, Définitions (Pourquoi ne peut-on pas ‘naturaliser’ la rai-
son) (L’éclat, 1992), p. 41.
3. J. Piaget, “Introduction,” in J. Piaget, ed., Logique et connaissance
scientifique (Pleiade-Gallimard, 1967), p. 3.
4. E. Husserl, Formale und Transzendentale Logik, in Jahrbuch für
Philosophie und phänomenologische Forschung X (1929), §5.
5. Ibid., §9.
6. Ibid., §6.
7. Ibid., §27.
8. M. Heidegger, Kant und das Problem der Metaphysik (Klostermann,
1991), §24.
9. I. Kant, Kritik der reinen Vernunf, A158, B197; translation by V.
Politis, Critique of Pure Reason (Everyman’s Library, 1993).
10. I. Kant, Introduction to Logic, T. K. Abbott, ed. (New York, 1963),
Chap. VII.
11. I. Kant, Prolegomena zu einer jeden künftigen Metaphysik, die als Wis-
senschaft wird auftreten können, op. cit. [1], §10.
12. I. Kant, Introduction to Logic, op. cit. [10], Chap. I.
13. I. Kant, Prolegomena zu einer jeden künftigen Metaphysik, die als Wis-
senschaft wird auftreten können, op. cit. [10], §13.
14. I. Kant, Introduction to Logic, op. cit. [10], Chap. I.
15. G. G. Granger, Formes, opérations, objets (Vrin, 1994), p. 75.
16. L. Wittgenstein, Tractatus logico-philosophicus (Routledge & Kegan
Paul, 1963), 6.13.
34 M. BITBOL

17. J. Piaget “Epistémologie de la logique,” in J. Piaget, ed., Logique et


connaissance scientifique, op. cit. [3], p. 385.
18. Ibid., p. 383; see J. Piaget, Introduction l’épistémologie génétique, 1/
La pensée mathématique (P.U.F., 1973.).
19. J. Piaget, “Epistémologie de la logique,” in J. Piaget, ed., Logique et
connaissance scientifique, op. cit. [3], p. 388.
20. Ibid., p. 397.
21. J. Habermas, Zur Logik der Sozialwissenschaften (Suhrkamp, 1982).
22. See, for example, J. Dewey, Logic: The theory of Inquiry (Holt, 1938).
23. J. Piaget, “Introduction,” in J. Piaget, ed., Logique et connaissance
scientifique, op. cit. [3], p. 4.
24. L. Wittgenstein, Tractatus logico-philosophicus, op. cit., 6.13.
25. Ibid., 6.124.
26. Ibid., 2.19.
27. Ibid., 2.171.
28. Ibid., 4.1.
29. L. Wittgenstein, Philosophical Investigations (Blackwell, 1958).
30. J. Bouveresse, Wittgenstein, la rime et la raison (Editions de Minuit,
1973), p. 67.
31. L. Wittgenstein, Philosophical Investigations, op. cit. [29], Chaps. II
and XII.
32. J. Bouveresse, La force de la règle (Editions de Minuit, 1987), p. 142.
33. H. Putnam, Définitions (Pourquoi ne peut-on pas ’naturaliser’ la rai-
son), op. cit. [2], p. 71.
34. L. Wittgenstein, On Certainty (Blackwell, 1969), §174.
35. Ibid., §94.
36. G. G. Granger, Formes, opérations, objets, op. cit. [15], p. 113.
37. E. Husserl, Formale und Transzendentale Logik, op. cit., 92.
38. M. Mugur-Schächter, “Space-time quantum probabilities, relativized
descriptions, and Popperian propensities I and II,” Found. Phys. 21,
1387 (1991); 22, 235 (1992).
39. E. Husserl, Erfahrung und Urteil (Glaassen & Goverts, 1954), §47.
40. Ibid., §8.
41. M. Bitbol, Mécanique quantique: Une introduction philosophique
(Flammarion, 1996).
42. N. Bohr, Physique atomique et connaissance humaine, with Introduc-
tion and annotations by C. Chevalley (Folio-Gallimard, 1991). F. Lu-
rat, Niels Bohr (Criterion, 1990).
43. P. Heelan “Quantum and classical logic: Their respective roles,” Syn-
these 21, 2-33 (1970). M. Bitbol, op cit. [41].
44. S. Watanabe, “The algebra of observation,” Suppl. Prog. Theor. Phys.
FORMALIZED EPISTEMOLOGY, LOGIC, AND GRAMMAR 35

37 and 38, 350-367 (1966).


45. M. Mugur-Schächter, “From quantum mechanics to universal struc-
tures of conceptualization and feedback on quantum mechanics,”
Found. Phys. 23, 37 (1993).
46. E. Schrödinger, The Interpretation of Quantum Mechanics, edited and
with introduction by M. Bitbol (Ox Bow Press, 1995). M. Bitbol,
Schrödinger’s Philosophy of Quantum Mechanics (Kluwer Academic,
1996).
47. For a comparative analysis of these diverse approaches, see: M. Bitbol,
op. cit. [41]; “Quasi-réalisme et pensée physique,” Critique 564, 340-
361; L’aveuglante proximité du réel,” Critique 576, 359-383.
48. M. Serres, Les origines de la géométrie (Flammarion, 1993), p. 21.
49. Ibid., p. 27.
36
4
EPISTEMIC OPERATIONS AND FORMALIZED
EPISTEMOLOGY: CONTRIBUTION TO THE STUDY
OF THE ROLE OF EPISTEMIC OPERATIONS IN
SCIENTIFIC THEORIES

Michel Paty

Equipe REHSEIS
CNRS and Université Paris 7-Denis Diderot
Centre Javelot, 2 Place Jussieu
F-75251-Paris-Cedex 05, France

We ponder the kind of problems and perspectives of a “formalized episte-


mology”, by considering the advantages than one get from a concern with
the “formal”, with its structural orientation, that would favour compre-
hensive, unifying and synthetic, intelligibility. We confront this perspective
with that of the changes in knowledge, considering the relation between form
and meaning for knowledge contents, and examine the notion of “epistemic
operation” as instrumental for creating new forms, at the theoretical and
meta-theoretical levels. Actually, the notions of form, of formal and of object
are not independent of the problem of a subject that decides on conventions
and choices. “Epistemic operations” might suggest a link with “algorith-
mic functions” for knowledge statements, that themselves entail the risk of
reductionism in a naturalistic conception of representation.

Key words: changes, contents, epistemology, form, meaning, object, opera-


tion.

1. INTRODUCTION

The expression “formalized epistemology” admits of a whole variety of pos-


sible definitions, thus staying widely open to interpretations.1 Taken intu-

1
The following reflections have benefited from exchanges and discussions inside the
working group entitled “Centre pour la Synthese d’une Epistémologie Formalisée (CeSEF,
Paris), animated by Mioara Mugur-Schächter. I acknowledge the friends and colleagues

37
38 M. PATY

itively, it is perceived to point toward a particular interest in formal repre-


sentations, in any given area of knowledge, and in the connection of such
representations with scope and meaning, and with the relationships between
different fields of knowledge. In what follows, we much less intend to reach
a precise definition of what a formalized epistemology should or could be,
than to focus on some problems characteristic of an investigation of the
type suggested by such a name and to stress both its interest and its lim-
its. Therefore, from the start, we tolerate vagueness in the definition of
the concept, in order to avoid an a priori confinement inside a too-narrow
and artificial set of boundaries: we mainly aim at exploring what the label
“formalized epistemology” could reasonably point toward.
Let us begin by specifying our project still more. Three quarters
of a century ago, Ernst Cassirer tried to establish a “morphology of the
mind”. First, in his beautiful work Substance and Function [13] he exposed
an inquiry into “the structure of thought” as manifested in mathematics and
natural sciences. Then, under the global title The Philosophy of Symbolic
Forms [15], he dedicated three new volumes to an analysis of language, of
mythical thought, and of the phenomenology of knowledge.2 In what follows,
Cassirer’s basic idea that any area of human thought—from its “infrastruc-
ture” to the “architectonic organization” of superstructures that constitute
the sciences 3 —is expressed via symbolic forms, while keeping oriented to-
wards the real world, will remain for us a fundamental source of inspiration.4
Now, questioning form and formal problems when one deals with exact sci-
ences, in general involves emphasizing some current aspects of the present
scientific theories, with an endeavour to bring forth their characteristics in
this respect. We do not want, however, to confine ourselves to static struc-
tural features. Our interest, especially from a comparative viewpoint, is also
directed toward the movement, the streams by which the formal aspects are
brought about, and is strongly impulsed by questions as to how in future
other formal aspects could possibly settle in. If we restricted ourselves to
examination of only the formalization of recent knowledges, we could “lose
the prey for the shade” retaining merely a schematic, a “logically” recon-

from this group for the rich debates we had together. A first version in French of the present
work has already been published: “Opérations épistémiques et épistémologie formelle.
Contribution à l’étude des opérations épistémiques dans les théories scientifiques”, Prin-
cipia (Florianopolis, Br.) 3 (2, December), 257-306 (1999).
2
To each one of these subjects, he dedicated a whole volume of The Philosophy of
Symbolic Forms.
3
[15], Vol. 3, pp. 13-14.
4
This idea has already oriented us in other studies, in particular in our La matière
dérobée (The Stealing of Matter), dealing with the conceptions of contemporary physics
[64].
EPISTEMIC OPERATIONS AND FORMALIZED EPISTEMOLOGY 39

structed view of the sciences; whereas the living reality of our knowledges,
their contents, their substance which generates the forms, regularly undergo
changes. This “life of forms” of which the arts are so explicitly aware,5 obvi-
ously has to be a constitutive dimension of also any attempt at a formalized
epistemology. So, though in the present approach our attention will mainly
focus on the forms themselves—on the “logic of forms”, on the analysis and
the meaning of forms—rather than on the historical circumstances in which
they emerged, we shall nevertheless also keep clearly in view this essential
dimension of continual change.
Therefore this essay, aiming to “problematize” the concept of a for-
malized epistemology, will begin with a reflection on the awareness of change.
We shall then continue with an examination of the notion of “epistemic oper-
ation”, regarded as being instrumental for the creation of new forms at both
the theoretical and the meta-theoretical levels. Then the specific features of
form and of the formalized will be examined, as well as their relations with
the contents of knowledge and with the notion of object, both considered
as depending on a subject’s decisions and on conventional choices. We shall
conclude with questions concerning the link between “epistemic operations”
and the possibility of an algorithmic representation of knowledge and of its
generation, which will lead us to emphasize how a naturalistic conception
about it entails a risk of reductionism.

2. EPISTEMOLOGY AND AWARENESS OF CHANGE

As remarked above, epistemology is quite essentially tied with change. Ein-


stein has pointed out that progress in physics leads to theoretical represen-
tations which are increasingly distant from our immediate apprehension of
reality. This is true for also most other types of science. Only, perhaps,
disciplines of which the mode of expression is narration—the characteristic
case of this kind is history—require a type of intelligibility that remains in
a direct and close contact with subjective impressions and immediate sen-
sations, from which the facts of a revolved past must be reconstituted and
reactivated. (But even this does not forbid a comprehension informed by
also more abstract and reflected elements, involving judgements and con-
structed assignation of meaning).
Those who still think that scientific knowledge aims to barely some-
how express “reality”—physical, or psychological, or social, etc.—should

5
See, for example, the classical work of Henti Focillon on the history of art, La vie
des formes [28]. Many titles of books on esthetics and on art bring in the word “form”,
in its common concrete meaning. And the question of the relation between form and
signification is obviously a central one in the domain of arts.
40 M. PATY

try to become genuinely aware of this incessantly increasing distance and


the corresponding mediations laid between representative thought and its
object. Even mathematics deal with “reality”, but a reality consisting of
idealized forms that are already placed at a large distance with respect to
our original intuitions about reality such as they emerged from the direct
sensorial experience. The numerous mediations which occurred between
these original intuitions and the idealized forms the scientific object consists
of are in a certain sense incorporated in this abstract object. Our idea of
reality incorporates something as traces of all our successive representations
of it. Indeed the successive states generated by the successive mediations
between our primary intuitions and the nowadays contents of our minds,
can be traced back through history: it is possible to identify them by recon-
sidering each particular knowledge such as it emerged, rooted in a definite
global culture where it was tied by definite relations to the other contempo-
rary knowledges. In each historical stage, the knowledges from that stage
and the global culture in which they were embedded, composed an organic
whole marked by specific types or norms of intelligibility. So the question
of the intelligibility of the world and of the nature of scientific knowledge,
is not separable from historical considerations. This, however, does by no
means entail a fundamental relativism that would deny or minimize the role
of scientific reasoning. One has to admit as just a matter of fact the ex-
istence of forms of knowledge and of modes of justifying reasoning which
differed from one another according to the historical and cultural contexts.
And this fact itself, in its turn, is also liable of a rational, scientific inves-
tigation, which nevertheless has to be posited to be ruled by norms able to
incorporate change, and to generate evolving forms.
These remarks, of which the implications concerning the nature of
scientific knowledge have been examined by us elsewhere [77], are expressed
here with the unique aim to stress the importance and extension of the ques-
tion of the relationship between a new knowledge and the tradition in which
it appears, hence, between present knowledge and future knowledge. This
question widely transcends our nowadays science. Though in this work, by
necessity, we are restricted to a domain of knowledge on which the nowadays
knowledge imposes an upper bound, it is fundamental to keep in mind that
the historical dimension, drawn by time, is irrepressibly mobile together with
its contents ; we have to stay fully aware that the knowledge inside which
we are now located, though scientific, nevertheless is neither immutable nor
“co-natural” to us, but is the result of elaborations which, indelibly, bear
the mark of historical circumstances; that, in their turn, our attempts at
achieving purely rational and formalized metarepresentations of our knowl-
edge, can only lead to results which are themselves constructions incorpo-
EPISTEMIC OPERATIONS AND FORMALIZED EPISTEMOLOGY 41

rating certain contingent and conventional features; that these contingent


and conventional features affect not only the objects of our representations
but also the modalities of representations, up to the very manner to conceive
them; and that all these changing characters at all these various levels, affect
equally—and possibly even more—all the other areas of knowledge and of
human experience which are not reducible to the scientific knowledge, such
as aesthetics or morals, these pillars of any culture which, albeit differently,
are also linked to the use of reason, while, like any knowledge, they bring
into play also the other functions of thought, like imagination and memory.
So, coming back to our nowadays knowledge, it should be clear that
in many areas the way in which we now conceive of what this or that phe-
nomenon is, to what sort of object it refers or what it admits of, differs
appreciably from preceding views in these respects, including some of those
which we ourselves have initially learned. However, although we might be
conscious of many among these changes, since we experience them directly,
we do not yet know clearly what they modify in our current manner of think-
ing and speaking, nor how [53-55]. This knowledge is indeed most difficult to
be gained while still staying inside the previous global traditions of thought.
We are living these changes before knowing how to think them. However,
in spite of all, it is unavoidable for science, and it is essential for philoso-
phy, to somehow undertake an effort for thinking them in a coherent way.
But is it possible, right now, to know something more about these ongoing
changes, and to formulate explicitly the new rules of thought required by
them implicitly?
A preliminary step would be to explore what can be hoped in this
direction. Since the present situation is not entirely new, since it has prece-
dents from certain points of view, we can try to draw lessons from the past.
Also, focusing attention to just the changes which are now occur-
ring, can be instructive even without necessarily understanding them, nor -a
fortiori—being immediately able to forge a new global view able to encom-
pass them. Our ambition, even if it is kept modest, might nevertheless be
fertile; while if it is too big at the start it might generate merely illusion.
Another, more constructive step, would consist in explicating direc-
tions inside the already known evolutions of thought. Here we shall evoke
those which took place inside physics, but similar examples can be found
in biology, geology, cosmology, mathematics, as well as in the social and
human sciences. For instance, the theory of relativity (special and general)
has obliged to rebuild the notions of space and time which were considered
to be the most obvious and stable ones and were the deepest rooted in our
cognitive structures. It has also brought forth the necessity to re-examine
the concept of physical theory itself, namely to regard it more and more
42 M. PATY

as a symbolic, conceptual and formal construction which nevertheless perti-


nently represents the real world [64,68]. Thereby it has enabled us to acquire
a more appropriate perception of the very peculiar role played by mathemat-
ics in this construction. We can thus conclude that mathematics manifest
an abruptlyincreasing importance in our representations of “reality”. This
conclusion yields a dimension where past evolutions can be pertinently em-
bedded, thus clarifying certain historical facts. Indeed, since already a long
time it became obvious that “the space-time continuum” is just an abstract
entity elaborated by thought, of which the justification cannot be regarded
to consist exclusively of evidence provided by intuition, but also—mainly
perhaps—of the operational power of this construct: the use of differential
equations as well as the basic concept of field of forces are based on this con-
struct, and this triggered accurate descriptions and explanations of a huge
amount of phenomena. But the operational power of the concept of space-
time continuum is fundamentally tied with its mathematical representation,
which strongly guided our most basic schemes of intelligibility throughout
long, progressive elaborations [71].
Similar remarks hold concerning also other conceptual, symbolic con-
structs, like material points, forces, etc., which, strictly speaking, are de-
void of a genuinely “real” counterpart. Newton’s mechanic of material
points, while it introduced the space-time continuum, also brought into evi-
dence the general necessity and outstanding usefulness of idea-like (“idéels”)
mathematical constructs founded on abstractions drawn from—supposed—
realities. Indeed most among the constructs of this sort were first introduced
inside the framework of newtonian mechanics, with its general principles and
laws, its own inner relationships and basic concepts. But later they have
constantly and progressively been reformulated, completed, and generalized
(to complex solid bodies, to fluids, to gravitational attractions between more
than two bodies, etc.). And, whatever be their still unclear relations with
what we call physical reality, such constructs proved so extremely fertile in
the hands of physicists and mathematicians that one feels strongly inclined
to conclude that, though nowadays they have become so very remote from
the real in the sense of our intuitive perceptions, nevertheless they somehow
exceed the schematic character of a mere mathematically convenient “ide-
ality”, that they somehow have definitively incorporated something drawn
from the “real”. So the direction of increasing mathematization seems to
belong organically to the past-and-future history of scientific knowledge.
But it is quite noteworthy that this pertinence and viability with
respect to the physical reality, of certain abstract constructs from mathe-
matical physics, owes also much to a deep epistemological concern of major
EPISTEMIC OPERATIONS AND FORMALIZED EPISTEMOLOGY 43

scientists ,6 who throughout centuries kept discussing the conditions of va-


lidity and the limits of applicability of the mathematical representations of
the basic concepts involved by physics. Which, again, illustrates the inti-
mate relation between change and epistemology.
As for the changes which quantum physics introduced in the con-
ceptualization of physical reality, these, from both the viewpoint of the
physical meaning of the theory, and the viewpoint of its epistemological im-
pact, are still far from having been fully evaluated. Very soon, the quantum
physicists themselves have explicitly tried to extract from this so peculiar
discipline, a general methodology for physical theories, and a philosophy of
knowledge. But precisely the concern to insure for this new and remarkably
fertile scientific theory, a legitimacy and an impact that seemed in danger
of being denied, fixed—prematurely, and in a rigid way—fictitious limits of
interpretability. The conceptual difficulties, pertaining to the physical ar-
gumentation as well to the theoretical problematic and the epistemological
analyses, were buried under the automatic answers of an ad hoc “positivis-
tic” philosophy. It is quite true that, in the light of the knowledge pro-
duced by the exploration of the new phenomenal area, it was necessary to
reconsider attentively basic categories—like causality or determinism, ob-
servation, object, objectivity—of which the function had seemed before to
be definitively understood. But at the same time many other concepts—like
those of state of a microsystem, of magnitudes or descriptional quantities
characterizing such a state, or the significance of the probability linked with
a state—concepts which possibly could come out to hold, in the questions of
interpretation, a role no less fundamental than the classical categories men-
tioned above, were totally omitted in the epistemological and philosophical
discussion. The quantum formalism introduced mathematical definitions
for concepts of which the relationship with a physical reference was left
obscure, while the very relevance of specifying such relationships was de-
nied and banished from the start on, on the basis of alleged philosophical
reasons [83,81,86]. In this way the intelligibility of quantum physics has
been artificially immobilized in an incipient stage. This is illustrated by the
recent debates on the meaning of what is called local non-separability (of
genetically tied quantum microstates) [6,63,27,110,18] and by the subsist-
ing doubts concerning the theoretical status and the physical meaning of
the “principle of reduction of the wave packet” (the problem of quantum
measurements) [119]): these debates point clearly toward important zones
of obscurity that persist in the question of the significance to be assigned to
the quantum mechanical formalism.
This situation is clearly unsatisfactory. Indeed the quantum the-
6
From d’Alembert to Mach, Boltzmann, Einstein, etc.
44 M. PATY

ory is not restricted to fundamental quantum mechanics, it extends to the


quantum field theories which today possess a considerable importance, and
which, in the long range, will manifest their own conceptual and physical
implications, certainly related to those of fundamental quantum mechanics.7
Therefore, when we speak of the general lessons to be drawn from quantum
physics, we do not mean what is currently called “the philosophy of quantum
mechanics”, which refers mainly to Bohr’s concept of “complementarity” or
to its variants according to others physicists, insofar that they put theoret-
ical and conceptual criticism under the dependence of a philosophy of the
observation. Much more generally, we mean a global understanding of all the
theoretical, epistemological and philosophical questions raised by the new
knowledge concerning quantum phenomena, which, with the relativistic ap-
proaches, constitutes one of the two most important corpuses of knowledge
from modern physics [82,83,81,86,85]. So the representations of quantum
phenomena have to be thoroughly studied from an epistemological point of
view: this is a main task of a nowadays formalized epistemology.
Various other contemporary physical theories have also surprising
implications concerning relationships that were usually considered to be
unquestionable, but which now require a rigorous critical reconsideration.
Consider for instance the one-to-one link posited between determinism and
prediction, which lasted unchallenged such a long time and now is found to
be inconsistent with the modern theory of nonlinear dynamical systems and
of the “chaotic” phenomena manifested by these: the dogma of predictibility
in terms of space-time coordinates, of the individual trajectory of a moving
body, looses importance, while other global descriptors, like the strange at-
tractors, are pertinently defined [108,109]. Many other such epistemological
re-evaluations are strongly needed today if one wants to grasp the whole of
our present knowledges accordingly to a wider, global view endowed with
inner consistency and offering a deeper intelligibility. Which possibly—
probably—would also entail reformulations. Obviously, an integration of
the kind alluded to above, could be achieved only by precise, “differential”
analyses of each of the already acquired knowledges. But these, if achieved
and then confronted with history, might surprisingly bring forth general
structural lines concerning some definite characters: in this sense, history
and new knowledge can be brought to work together for the elaboration of
a more integrated and deeper epistemology. Of course, regularities of the
mentioned kind cannot be perceived at a first glance, nor with full generality
from the start on. Passage of time, with the sedimentation entailed by it, is
necessary for the regularities to draw attention upon them. Any regularity,
when it first becomes perceptible, is perceived as a mere coincidence, while
7
[64], Chap. 8.
EPISTEMIC OPERATIONS AND FORMALIZED EPISTEMOLOGY 45

coincidences and analogies are—rightly—felt to be too weak to be assigned


thorough consideration.8
One may wonder, however, whether it would not be possible to imag-
ine a point of view or an axis for research able to constantly yield a global
perspective on the facts and problems with which the regional, specific epis-
temologies of the various theories are confronted: a perspective that would
keep encompassing their wealth and variety while also giving at the same
time an insight into the main structural currents that organize them into
one whole according to some common, unifying, synthetic intelligibility. Get-
ting at such a point of view might allow to better control the unavoidable
changes which occur inside any sort of knowledge, to discern the main di-
rections generated by them, to set up a reasoned inventory of these, and
perhaps to anticipate other ones.
Such an anticipation, founded on a systematic search for regularities
or trends, is not unthinkable. Indeed, one can investigate concerning the
reason of similarities, analogies and convergences observed in the transfor-
mations occurred inside totally different regions of knowledge, and to strive
connecting them to morphological or functional features which, like a pre-
set conditioner, enveloped them all from the start on and was doomed to
eventually manifest itself. If this were true, previous systematic examina-
tions would possibly have been able to become aware of such acting pre-set
conditioners, without having had to wait for the whole variety of cognitive
events to occur in a contingent way. Such considerations suggest a kind of
meta-epistemology, to be invented, if only it is possible. This, with respect
to the regional epistemologies of the various definite theoretical representa-
tions, would yield a kind of an analogue of what, in Hermann Minkowski’s
eyes, would have been an a priori mathematical theory of space-time if it
had been developed before the physical theory of special relativity [51,52].
Let us note that hypotheses of this kind seem to involve a wholly rational
conception on the emergence and evolution of the theories and knowledges,
and of their links with the meta-theories liable to frame them. Which, how-
ever, would require specification of what is to be understood by a “rational”
conception of the mentioned kind, compatible with some allowance left for
also invention. For, if not, could we conceive of invention to be merely the
unpragmatic apprehension of our accessing to progresses in knowledge which
“in fact” were determined with an absolute necessity?
Anyhow, without prejudging on the answers to all these questions,
the preceding considerations incite to bestow particular attention upon that
which, in the stage of mutual readjustments of previously established knowl-
edges and even while the elaboration of new knowledges, comes up in the
8
On analogy, see Paty [88].
46 M. PATY

kinds of the “operative” and the “formal”. What is to be understood by


these terms, however, has to be closer examined before trying to set land-
marks for a reflection on what can possibly be expected from an approach
founded on the concepts they refer to.

3. EPISTEMIC OPERATIONS

We shall call epistemic operation an act of thought (or a series of such


acts) by which a body of knowledge is constituted, no matter whether yes
or not the nature of this act is consciously perceived while the process of
generation of knowledge goes on: this nature can be recognized in only
a further stage, when the contents and procedures from a given area of
knowledge are examined.
A simple example of epistemic operation can be found among the cur-
rent methods of contemporary physics, namely in the search for invariants
while building a definite physical theory: Lorentz-invariants are selected in-
side the framework of special relativity; what stays invariant under a given
“gauge” symmetry posited to be fundamental for the representation of some
given kind of dynamic interactions between “quantum” particles, is selected
as a relevant quantity in the theory of these interactions; etc. This prac-
tice has become usual in physics since the advent of general relativity and
of quantum physics, in the years 1920-1930. But its origin can be found
in the mmoire composed in 1905 by Henri Poincaré on the dynamic of the
electron [92] (La dynamique de l’electron), at the time when this author was
planning to build a theory of gravitation modified with respect to that of
Newton by imposing a condition of “covariance”, or invariance under the
“Lorentz transformations” of space and time coordinates, of the equations
expressing laws, which means subjection to the principle of (special) relativ-
ity. This practice in theoretical physics is related with the importance gained
since then by the connected notions of group of transformation, symmetry
and invariance, as defined inside the corresponding mathematical theory.
This importance has been fully understood with Einstein’s theory of gen-
eral relativity. It received a first formal systematization with the theorem of
Emmy Noether [58,59]. It has later guided the elaboration of quantum me-
chanics and then of the quantum field theory, up to the recent developments
concerning the fundamental interaction fields obeying gauge symmetry (in-
variance).
This sort of epistemic operation has considerably modified the con-
cept of physical theory as well as the everyday practice of the theorists.
Once one admits it to be justified, it is easy to formulate; but an a pri-
ori justification by use of some simple explanatory scheme would distort
EPISTEMIC OPERATIONS AND FORMALIZED EPISTEMOLOGY 47

or ignore the “facts that resist” of historical reality. After its “invention”,
an invariant usually is so well understood that no justification seems to be
needed. But this does not entail that we can regard it as a natural a priori
evidence. If in our retrospective look the invariants we make use of nowa-
days appear to be endowed with such obviousness, this is so because we
consider them from inside a conceptual universe where they already work
as a reference, in consequence of a radical reorganization of our knowledges
and of the methods of theoretical physics. But the concept of invariant itself
clearly possesses a historical origin that can be traced back to structural and
conceptual changes occurred in physics at the beginning of XXth century.
This concept appeared—it has been invented—in a conceptual world still
strongly marked by conceptions and practices radically different from the
nowadays ones, and in circumstances the study of which pertains to history,
and more specifically to historical epistemology.
This example, by its relative simplicity, permits to clearly perceive
that, in our study of epistemic operations, we should distinguish two levels:
a first level of study of the use of this or that epistemic operation, inside the
scientific work (in our example, the use of invariants for formulating laws
and physical theories); and a second level of study of this same operation
but considered to a “second degree“, as stemming from procedures involv-
ing historical aspects which have triggered its invention (invariants, in our
case). This second level is that of the constitution of epistemic operations,
of their elaboration, and it is not reducible to only descriptions of the oper-
ations themselves: this is connected with the question of the genesis of new
conceptions, of creation of novelty in science, and more generally, with the
question of the emergence of new forms in the cognitive thinking.
One can also consider as an epistemic operation the fact of reasoning
inside the framework of some given logical structure and some category of
thought that inform our “interpretations”, our manner to assign meaning to
the concepts and the theoretical statements we make use of. For instance,
concerning the propositions from a physical theory, the notion of causality,
some specific way of understanding determinism, the meaning ascribed to
the concept of probability. Interpretations in this sense act on the manner
in which a problem is processed, this manner being possibly common to re-
searchers and specialists from a given period, or else, differing according to
the individuals or the schools of thought. But they concern even more the
way in which knowledges are understood and justified; this way, moreover,
acts upon the dynamic of thinking and upon the decision to continue on a
given direction of investigation, or not. For instance, the notion of “theo-
retical completeness”, raised by general relativity and by quantum physics,
suggests a program of research in view of modified theories, program that is
48 M. PATY

accepted or not according to the position adopted with respect to the con-
sidered problem [65,84]. None of the two mentioned theories is complete in
the “strong” sense of a “self-generation” of its own objects, which is indeed
the sense involved by the present attempts at unifying the field theories, so
the question which persists is to know whether they are complete in some
weaker sense (are they sufficient for determining the properties that can be
assigned to their objects?):9 this, for Einstein, was a fundamental condi-
tion required for trying to accomplish a unified description of the physical
object-entities.
New results obtained in a given science might challenge epistemic
operations which otherwise would have been considered to be definitively
obvious. Such has been the case for the concept of causality that has been
modified by the theory of special relativity which, by obliging to distin-
guish between the space-like and the time-like regions of the light cone of
an event,10 has entailed changes in our conception of the relationship be-
tween cause and effect. All the regions of the space-time diagram are not
equivalent: if the time-like region is physical, the space-like region is non
physical (there is no possible causal relationship between its hyperpoints).
Henri Poincaré himself, with regard to his own ideas, asserted this concern-
ing Minkowski’s spacetime.11
On the other hand, the quantum mechanical concept of “probability
amplitude” entails a modification of the idea of probabilities as formed in the
classical physical theory. Probability is a mathematical concept. When ap-
plied in physics, it is generally identified with the limit of convergent relative
frequencies of events, according to the law of great numbers. The construc-
tion of quantum theory makes use of it in a way which is indirect. Namely,
the theoretical probabilities postulated concerning the quantum mechanical
states are calculated from the “probability amplitudes” or state functions or
vectors belonging to the Hilbert space of the microsystem which represent
the considered microstates of this system, with operators taken as the quan-
tum mechanical variables. These theoretical probabilities are afterward put
in relation with the experimental relative frequencies of physical observed
events. But since a theory defines the meaning of the quantities it makes
use of (via the relationships between these) independently of any experiment

9
This was the essence of the “EPR argument” [26]; see Paty [72,89].
10
The light cone, defined by the equation x2 + y 2 + z 2 − c2 t2 = 0 , determines an inner
time-like region, such as x2 + y 2 + z 2 < c2 t2 , and an external space-like region, such as
x2 + y 2 + z 2 > c2 t2 . The former is the region of causal relationships between spacetime
points, the latter is that of a-causal relationships (“nonphysical” region).
11
Poincaré [95]. Cf. Paty [73,74]. This was shortly after Paul Langevin had discussed
in philosophical meetings, in Poincaré’s presence, the physical implications of the new
relativist conceptions on causality (Langevin [46,47]; cf. Paty [73,84].
EPISTEMIC OPERATIONS AND FORMALIZED EPISTEMOLOGY 49

(the experiments providing only knowledge of particular values occurring for


these quantities), the theoretical quantum mechanical probabilities possess
a specific theoretical meaning, and this is not reducible to the classical fre-
quency meaning of the probability of an event (think of the probability of
events involved in the self-interference for one single photon); it posseses a
meaning which differs from that of the probability of the result of a cast of
a die, as Dirac remarked already in 1930 [22,81,86].12 All the well-known
controversies on this subject are nothing more, in fact, than intuitive sub-
stitutes for a crucial conceptual breakthrough, namely that which led from
the classical concept of a descriptional quantity endowed with numerical val-
ues that can be conceived to pre-exist, and the quantum mechanical concept
of a quantity, which is more complex, rejecting the mentioned conception
[85,86,87].
The question whether the epistemic operations can be identified with
certain algorithms will be discussed later. First we shall have to make clearer
what we mean by the term “formalized” from the expression “formalized
epistemology”: We would like to show that, and how, this meaning largely
transcends that usually assigned to the word algorithm. We shall also con-
sider more thoroughly the concept of “object”, which will permit us to spec-
ify more narrowly the purpose of a formalized epistemology, with respect to
general features of the epistemic operations.

4. VARIATIONS ON THE THEME OF A FORMALIZED


EPISTEMOLOGY

In the classical sense, formal —thus also formalized—somehow opposes to


material, as for example in the case of Aristotle’s formal cause (bearing on
an idea or an essence), or as in the antagonistic couple form-matter, or else,
in the most current meaning of the word “formalism” (“purely formal” is
meant as devoid of a “real” content in the sense of a “material” one)13 .
This opposition, exploited by scholastics, refers to the shape of the relations
existing between the elements of a cognitive operation, abstraction being
made of the “matter” or meaning (reference) associated to these elements.
With this meaning the terms form applies, for instance, to expressions such
as formal relationship in algebra, which point toward connections that stay
valid for any numbers so that literal symbols can be used instead of this or

12
We leave aside here the various interpretations that have been proposed for probability
in this context, from Werner Heisenberg’s “potentialities” to Karl Popper’s “propensities”
[99,101]; these notions, as interesting as they might be for pointing toward certain prob-
lems, nevertheless are only “intuitive” and vague.
13
Lalande [45]: articles Cause, Formalisme, Forme, Formel.
50 M. PATY

that definite number.14 The concept of form is also part of the terminology
concerning natural laws (the form of a law), while in philosophy, for example
the kantian philosophy, the laws of thought involve pure forms of the sensible
intuition, a priori forms of the sensibility (time and space), as well as forms
of intelligibility (categories) or forms of reason (ideas). In a different, more
recent sense introduced in the Gestalttheorie, the theory of structure in
psychology, the word form points toward what obliges to consider an element
as part of an organized totality, as participating of the structure and its
structural laws.
Taking into account this whole variety of meanings, we can now get
closer to a significance able to be directly useful to us in an attempt to
reach the fundamental features of a gnoseological approach of the contents
of knowledge.
Gilles-G. Granger speaks of “formal contents” in the case of math-
ematics, to be distinguished from the empirical contents involved in the
natural sciences while at the same time bringing them nearer to these last
ones: the mathematical forms are not empty forms, they also possess certain
contents, namely those expressed by relationships that are not reducible to
tautologies like the logical axioms, nor to mere well-formed expressions in
the sense of symbolic logic [33,34]. Indeed the “formal” in this meaning, i.e.,
of formalized contents, cannot be identified to the “purely” logical, which
is exclusively formal by construction, since, by definition, it is not opposed
to content. In what follows we intent precisely to clarify the relationship
between the formal and the corresponding content in the case of a formal-
ization of a content (aspects of this relationship will appear in the discussion
on the concept of object). Thereby the opposition between content and form
will be transcended.
“Formalized” can be opposed to what is a particular empirical de-
scription at a phenomenal level, even if this description has been obtained
inside a natural science, via a theoretical approach. Consider mathematical
physics and theoretical physics, between which we distinguish. What math-
ematical physics is mainly interested in, are the formal relations between the
mathematical quantities posited to qualify physical entities (objects or phe-
nomena). This way of being formal is of the same kind as that which occurs
in mathematics; while in theoretical physics one is mainly concerned with
the physical contents pointed toward by the involved mathematical relations.
Mathematical physics appears in various respects as a formalized approach
if it is compared with experimental physics regarded as the source of con-
tents; and the same assertion holds even if it is compared with theoretical
14
In ancient writings (like the scholastic ones) “formal” calls forth “actual”, while in
the mathematical or logical sense it calls forth the idea of “general”.
EPISTEMIC OPERATIONS AND FORMALIZED EPISTEMOLOGY 51

physics. However, rather often theoretical physics converges with mathemat-


ical physics, whereby the referents of these two denominations are brought
to a momentary identification (like in the case of the analytical mechanics as
exposed by Joseph Louis Lagrange and Rowland Hamilton [70,78]; or Her-
mann Minkowski’s formulation of special relativity [51,52,116]; or Einstein’s
theory of general relativity, or certain presentations of quantum mechanics
(in particular those of Hermann Weyl and John von Neumann [17,57]), or
many developments of gauge field-theories (consider the works of Yang and
Mills), up to the most recent researches on quantum gravitation).15 But the
above distinction between mathematical physics and theoretical physics is
only relative, and the periodic identifications between a piece of mathemat-
ical physics and the corresponding piece of theoretical physics prove that,
as highly formalized a piece of mathematical physics might be with respect
to the representations of natural phenomena achieved in theoretical physics,
the mathematical relationships involved by such a representation never be-
come strictly alien to the physical contents of phenomena: only, at certain
privileged limits, the determination of the form of the mathematical rela-
tions that express these physical contents is so unambiguous, so achieved,
that it is exactly the same inside both theoretical physics and mathematical
physics, and then this superposition, by suppression of any comparability,
generates an illusion of absence of physical content, so of “purely” formal.
Coming back to the previously mentioned case of invariants, these, at
a first sight, might be perceived by thought as purely formal relationships,
but in fact they express general and fundamental properties of the physical
systems and of the quantities by the help of which these are described in the
considered physical theory. So, far from being external and superficial, the
invariants express physical contents, they are bearers of meaning, privileged
bearers of certain specified sorts of meaning, maybe the only possible bear-
ers for those sorts of meaning. It is in this way that Poincaré considered
the “mathematical analogies,” 16 and Einstein considered the “formal analo-
gies,” 17 which amount to the same thing (for mathematics is the “formal”
of the physicist). Poincaré and, some time later, Einstein did not hesitate to
speak of a “heuristic of the mathematical formalism” that drives the “phys-
ical thinking”, precisely because this formalism, in the cases considered by
them, was impregnated with, and informed by the physical meanings which
it served to express.18
Although the nature and role of the mathematical formalization in

15
Notably, Ashtekhar [1]. Cf. Kouneiher [41].
16
Poincaré [91]. Cf. Paty [88].
17
Einstein [23]. Cf. Paty [68], Chap. 4, pp. 164-172.
18
Paty [68], Chap. 5.
52 M. PATY

mathematical physics draws already attention to the fact that formalization


by no means excludes content and meaning, this does not yet tell us what,
exactly, should be conceived to be the purpose of a “formalized epistemol-
ogy”.
Does one mean, for example, a study of the formal aspects of various
epistemological approaches, in the way in which certain art-critics study the
formal aspects of some given set of works from some given area of artis-
tic expression? As for instance in the case of the critical analyses of the
literary forms occurring in Baudelaire’s or in Mallarmé’s poetry? Or oth-
erwise, does one imagine a study of forms the aim of which be not exclu-
sively to describe or to characterize existing “styles” of writing or thinking,19
but which are furthermore intended to disclose new forms of expression, or
even—operationally—manners to generate new contents, new meanings by
concentrating attention upon the form, for example by imposing formal
constraints while achieving a work of some given charcter. This tendency
is frequent in contemporary art, in painting, in music, or in literature (con-
sider the exercises of the Oulipo group20 ) where poetical or literary writing
is submitted to formal constraints,21 which generates aesthetic innovations
and “effects of meaning”.
Consider what might be called an epistemological reflection: it also
deals in part with forms, and it also takes on forms itself, forms that depend
on the modalities of the practised approach and that are connected to ques-
tions of meaning, in a somehow more direct and compelling manner than
the aesthetic forms: a meaning which, even if it is conceived to pre-exist, is
not necessarily also known for that, and which could be tried to be brought
into light from under the facts and appearances. Such an epistemological
reflection would certainly not be reducible to a free creation of form, nor to
a creation of pure form. In any study, the object to be studied is in a certain
sense, at least to a certain extent, given in advance from outside the study,
and the study suffers constraints entailed by this object and tied with its
externality. In particular, in the case of any given sort of scientific knowl-
edge, mathematics included, the source of the constraints which restrict the
representation can indeed be mainly assigned to the object of the knowledge
to be elaborated. Now, will a similar situation manifest itself in the case of
this meta-study which we want to call a formalized epistemology, devoted to
scientific knowledge as a whole regarded as the “object” of study? (We let
19
On the notion of style in science, see Granger [31], Paty [66], Chap. 4, [68], Chap. 1.
20
Ouvroir de littérature potentielle (Opening device of potential litterature) created in
1960 and animated in particular by Italo Calvino, François Le Lionnais, Georges Pérec,
Raymond Queneau, Jacques Roubaud (Oulipo [61]).
21
See, e.g., La disparition, by Georges Pérec (Pérec [89]), a novel written without using
the letter e, a vowel yet so ubiquitous).
EPISTEMIC OPERATIONS AND FORMALIZED EPISTEMOLOGY 53

aside for the moment the question of formal constraints possibly imposed
by the gnoseological structures from the subject’s mind).
A “formalized epistemology” should be, so to say, in close connivance
with its object. This excludes any formalisms that would be more or less
alien to this object, that is, to scientific descriptions of domains of real
facts. It also excludes any identification of the desired formalized epistemol-
ogy itself, with a formalism that would not be directly relevant to its object
(like arbitrary algorithms or exercises dealing with arbitrary models). So a
formalized epistemology might focus either on forms that characterize for-
malizations of pieces of scientific knowledge, or upon the operations by which
these forms have been established (which we called epistemic operations).
If this is agreed upon, the margin still remains large for defining now the
manner, or manners, of realizing the aim just sketched out; which, inciden-
tally, is an advantage: it thus remains possible to choose the manner which
a priori seems to be the most adequate, and so the most fruitful one.
On the other hand, formalized does not mean quantified. The mathe-
matical representation of physical phenomena illustrates this via the distinc-
tion entailed by it between the qualitative consequences of the mathematiza-
tion (in the modern sense, not in that of Aristotle and of the scolastics) like
the mutual disposition of the defined quantities and of the levels of order
of magnitude that reflect the involved conceptual constructs, the ideas, and
on the other hand the quantified, the measures, of which the expression is
numerical.22
Furthermore, formalized does not mean schematic, in the same sense
in which describing the form of an animal does not mean to reduce it to
its skeleton: it concerns the fact and the manner of taking on a form. And
the question of the form in which scientific knowledge in general and the
particular sciences manifest themselves, either when already constituted or
while coming into being, suggests that form has to do with intelligibility.
Indeed what sort of intelligibility is offered by a given scientific knowledge
via the form acquired by it?
Also, though between form and structure there exists a narrow link,
it nevertheless is necessary to distinguish them from one another, form being
more global than structure.23 The form expresses the structure, but only
partly; it also expresses non structural characters but which must be com-
patible with the structure. On the other hand, a same piece of knowledge
may dress on various forms and one might wonder whether these correspond

22
See Paty [69].
23
On structure in science, see, e.g., Stegmuller [112,113]; and often less defined. On
structure for history, cf. Foucault [29], Veyne [114,115]; and, for anthropological represen-
tations, cf. Levi-Strauss [49].
54 M. PATY

to different intelligibilities. This question, in fact, linked to that of “inter-


pretations”, appears as a quite fundamental one: one could see there one of
the key-articulations of a “formalized epistemology”.
“Epistemology of formalized knowledges”, “epistemology of form and
of formalization”, “formalized epistemology”, . . . A scientific meta-approach
concerning the function of form in knowledge can also be seen as a method-
ology on the basis of its heuristic-oriented preferences. It remains to know
to what an extent it can pretend to generate norms for conceptualizing and
for reasoning, norms which would enable to reproduce or to anticipate in-
ventions (that are precisely inventions of forms): this is not obvious a priori
given what is known up to now concerning understanding and creation of
concepts. In certain sense reconsidering a concept amounts to recreating it,
and this happens inside the unity of an isolated, unique, subjective mind
(subjectivity being regarded to be the place where reasoning can occur).
Reasoning, creation of new conceptual forms, does not operate exclusively
with countable rules that lend themselves to classification according to some
typology; it goes on inside a consciousness which, throughout the processes
of understanding and constructing, mobilizes many other instances besides
those consisting of the identified “elements” of the problem to be solved. By
“other instances” we do not mean the psychological or sociological ones of
which, in a first stage at least, abstraction has to be made in the present
investigation, we mean factors which in general keep implicit, unconscious,
but which nevertheless play their part in the global economy of our acts of
conceptualization and reasoning.24
The preceding remarks lead us to try to situate the concepts of a
formalized epistemology and of epistemic operations, with respect to judge-
ments and to the decisions, choices, occurring inside the mind of the subject
of knowledge regarded as an epistemic subject. Which leaves then to be spec-
ified what remains of this subject—the unique place for intelligibility—that
can be assigned to “objective” knowledge which, as a matter of principle,
is required to abstract away any subjective singularity, keeping exclusively
the “general subjectivity” as expressed by operation, process and content
“in any subject” or, as some have told, “without subject”. However, inso-
far that knowledge is produced, evaluated, communicated, by means of acts
of creation and of judgement, the existence of subjects as the loci of such
acts is unavoidably necessary in an epistemology, even in a formalized one.
Without them knowledge would bear on contents devoid of intelligibility,
or at best endowed with an anonymous and abstract sort of intelligibility:
but one is in right to wonder whether contents in this sense are genuinely
24
On scientific invention, see Hadamard [35]; concerning the rationality of this invention
as confessed by several scientists, see Paty [79].
EPISTEMIC OPERATIONS AND FORMALIZED EPISTEMOLOGY 55

conceivable: can a content of knowledge be merely schematic?


A subject is always present in filigree beneath certain elements which,
upstream or downstream of any knowledge, qualify the conditions, the
modalities and the effects of the epistemic actions that have produced that
knowledge. Here we restrict to only mentioning the presence of such ele-
ments. Their enumeration and study belong to epistemology in the general
sense, structural or historical: conditions of possibility,25 fields of rationality,
styles, programs, intelligibility, intuition . . . .26 In the present context, the
difference between formalized epistemology and epistemology in the general
sense is that, if the second one takes into account the above mentioned el-
ements as objects for study, the first one takes them into account only as
a given conditioning datum, which it wants to transcend or, more exactly,
with respect to which it situates itself in order to bring forth “structural
invariants”.

5. OBJECT AND CONVENTION

As epistemology in general does, examining in particular epistemic opera-


tions leads to discern in the operations of knowledge a preliminary part, of
a conceptual pre-organization by the mind, which “prepares” an object of
knowledge, or rather, the conditions of its identification.27
All sciences, whether exact, natural or social ones, are nowadays con-
scious of the necessity of a critical re-consideration of the concept of object by
taking into account that an object is defined by separation from the subject
who introduces it, which entails a critique of also the concept of objectivity.
On the one hand, no object can be designated in the absence of a mental act
tied with an intention of the acting subject. On the other hand, any object
is defined by its distinction or separation from—both—the subject, and a
background against which it stands out. It is furthermore well-known that
science cannot be restricted, as for the conceptualisation of its objects, to
the characters assigned by common sense: in this respect again, quantum
physics offers particularly precise lessons, to which we refer without being
able to detail them here.28
There undoubtedly exists in science, especially since the belief in the
25
Kant [40].
26
Granger [31], Lakatos [44], Zahar [121], Chap. 1, Paty [64,65,66,68].
27
The term “preparation”, rightly stressed by M. Mugur-Schächter (op. cit.), comes
from quantum mechanics. But awareness of this rather universal pre-organizing procedure
(to which quantum mechanics has associated a peculiar flavour and an increased precision)
was present already in previous epistemological reflections (see, in particular, Margenau
[51]).
28
Paty [64,65,63,83].
56 M. PATY

uniqueness of geometry has dissolved, a remarkable trend toward a repre-


sentational transformation of the objects, in relashionships. This trend is
most definite in mathematics, but in physics also it is quite clear, supported
by the fact that physical theory takes on increasingly the form of a math-
ematical physics. Now, many have tried, or still try, to present this trend
as a “de-ontologization” of the sciences. However, by this trend, the con-
cept of object is by no means abolished in that of “mere” relationship. For
in order to deal with relationships, there must be some entities which are
related, say, some “elements”, even if the nature of these is undefined or
problematic. These—at least—are the “objects of the relationship”. And it
is indeed in this way that one tends to conceive of the concept object inside
mathematics and inside modern physics as well. It is also true that, while
the progress of increasingly abstract formalization of the physical theories
proceeds, the related “elements” themselves, in their turn, go over into new
relationships, partly at least. In a mathematical and in a physical theory
these “elements” can, and actually are, defined in and by the relationships
in which they partake (such is eminently the case in the quantum theory).

But, exactly insofar the system of such relationships is not entirely “trans-
parent” or “tautological”, it expresses a—structured—content that resists
total dissolution into the “exclusively relational”, whatever this might mean;
or, to specify it, meaning relations without anything being related.

This content which resists total dissolution in the relational and which is
what has to be known, possesses the attributes generally pointed toward by
the term “reality”. An example which at a first sight is simple is offered
by the system of integer numbers as organized by arithmetic, for at a closer
inspection this system brings forth the complexity of the involved relations,
with the type of “reality” underlying it. Though it is possible to generate
any integer by starting with the number 1—the unit—and by operating “ad-
ditions” of this unit, first to 1 itself and then, progressively, to the number
constructed in this way by preceding additions (and labelled 2,3, etc.), it
is not possible to know in advance all the properties of the numbers con-
structed in this way, for instance whether some given such number is prime
or not, what relationships with other numbers it does satisfy (think of Fer-
mat’s great theorem), etc. Note, however, that the generating relationship
(repetition of the elementary additions of one unit) implies not only num-
bers, but also opeation and number of operations in numbers, so that the
mode of generation looks simple in appearance. The related “elements” and
the relations between these form together one organic whole of form-with-
content wherefrom the content cannot be eliminated thereby loosing also
the form.
EPISTEMIC OPERATIONS AND FORMALIZED EPISTEMOLOGY 57

Physics also provides good examples. Consider the concept of field,


cleared by the special theory of relativity of the material support called ether:
all the physical features of what is called a field are defined by the mathe-
matical equations of that field. Or else, consider the non-distinguishability
of identical quantum systems (“particles”) which defines at the same time
the relationship and the object of this relationship; and also the symme-
tries of elementary particles that determine these particles-in-their-mutual-
relationships, via their interaction-fields (gauge fields).29
It might even be more adequate to say that the relations are what
becomes as concrete as something real. For if the transformation, always
possible, of given elements into others “more relational” ones makes the
“objectal” nature of these more and more relative, it nevertheless remains
that the relationship itself finally appears as endowed with all the charac-
ters of what calls an “object” endowed with an own consistency of a peculiar
nature. A nucleus of structured relationships closely interwoven is devoid
of none of the characteristic features of what an “object” means, once freed
from the “substances” of the ancient metaphysical doctrines. So the con-
cept of “object”—mathematical or physical (and, by extension, chemical,
biological)—looses its traditional relation with a directly ontological per-
spective. With regard to it the category of “being” is not deleted, but is
nevertheless removed at a certain distance by increasingly many relational
mediations. But the predicates of existence remain, because they are of a
non removable epistemological nature. So the ambiguities of the meaning
assigned to the word ontology cannot justify a suppression of the concept
of object. A representation or a theory is always representation or theory
of something: by definition, this “something”, the necessity of which is a
logical one, is the “object” that this theory designates and characterizes.
As soon as the contours of the representation or of the theory at
stake have been drawn, ipso facto its object has somehow been determined.
Obviously it is by an act of the thought that we have designated it as the
object to be described by that representation or theory. Such an act contains
a choice to separate this object from the “rest”, and it indicates a program
involving certain conventions which the object, in this sense, carries with
it. The conception on objectivity is correlatively affected by this state of
affairs: it is not only given (because thought is drifted along by the object),
but it is also decided (according to some norms or conditions) and built
as well (in relation with our choices in positing the object). The chosen
conventions depend upon the concepts and the theoretical system which
weave them together in order to describe the object. They are relative to
this system, and alternative conventions are thinkable, that do not bear on
29
Paty [82,81,86].
58 M. PATY

only the theory, but also on the totality of the elements of meanings for the
theory which insure for it a definite intelligibility, for which the criteria are
themselves partly meta-theoretical. It important to study and to specify in
precise terms the interplay of these two concepts, object versus convention,
which calls forth requirements that vary with the considered type of theory.
We add only a few more words concerning this point. Quantum
mechanics is often asserted to have eliminated the concept of object, at
least insofar as this concept is considered to be independent of any act of
observation and of conceptualization and to preexist to any such act. It is
true— and commonly admitted, even beyond quantum physics, as already
stated above—that the object has become object by an operation of the mind
which has separated it from the rest (or has prepared this the conditions of
this separation). But this “condition of possibility” of thinking the object
does not exhaust the descriptions which one can perform of it: in fact, it
only makes these possible by opening up the field for their realization.
In this respect one may consider that the specific problem of quantum
mechanics is that of the nature of the acts of thought and of the operations
required to get access to the description of the object to be studied, or, more
exactly, of its state. I have tried elsewhere to show (here I cannot detail on
this) that quantum systems and their states can be conceived in terms of
objects endowed with properties, at the cost of an extension of the meaning
assigned to the concept of physical quantity, beyond its usually accepted
meaning of a directly numerically valued quantity, and of a correlative ex-
tension of the concept of state function:30 in the practical understanding of
the physicists these extensions are already performed but they are not yet
explicitly admitted. Notice that already before the construction of quantum
physics one could have find of “objects” represented by abstract magnitudes
that had already lost the features usually assigned by common sense to any
magnitude (such has been the the case of the light wave, spreading out in the
whole space, or the field without the support of an ether, as evoked above).
Let us also remark that the concept of state function, already involved in
Hamilton’s mechanics, was suggested to him by optics with its principle of
minimum, which refers to another still more ancient origin that can be found
in both mechanics and optics, namely the principle of least action.
Let us retain at least what follows: it has been possible to build
representations that were said to correspond to a definite object, without
being in possession of a deep justification other than the mere fact that these
representations proved to be effective: this sufficed for asserting that a de-
scription of the object and of the phenomena related with it was available.
And the encountered epistemological difficulties always concerned the tools
30
Paty [81,86,87].
EPISTEMIC OPERATIONS AND FORMALIZED EPISTEMOLOGY 59

that were made use of, rather than the nature of the object of the considered
theory. In fact the conceptual-theoretical tools from a theory have always
been adjusted in view of the object they were intended to make us conceive
of: not to make us perceive it directly, but just to make us conceive of it via
the mediation of the theory constructed in order to represent it. In partic-
ular, the operational tool consisting of the quantum mechanical algorithms,
has been elaborated by a process of adequation to the aim of representing
coherently a certain system of objects, namely the world of quantum objects
and of phenomena tied with them. By the logic of their fabrication, the tool
and the elements of the quantum mechanical representation are made of the
same stuff. And the formalism of state functions defined in a Hilbert space
and of operators acting on these, which are aimed at representing states
of quantum systems, involves by construction the rules for making use of
it. However, inside their own physical domain of existence, the “objects”
toward which the state functions point, namely the states of quantum sys-
tems, do not need any more, in order to be thought and specifically referred
to as physical entities, to still be constantly referred to the abstract tools
which, in a certain sense, have led to bring them into being, and which also
permit to detect them (in terms of events coded in a language of eigenvalues
of eigenfunctions of quantum mechanical operations and of probabilities of
such events). But in fact they are systematically conceived of accordingly
to the way in which they have been designated, so in full agreement with
the quantum mechanical formalism.31 In a way, the well known problems
raised by the quantum mechanical formalism is not so much that of the
represented quantum objects, as that of the relation between this quantum
representation and the classical “mechanical” representation, adapted to the
involved experimental devices.32
These remarks suggest a process which in a certain sense is opposite
to that of a progressive “syntactization of semantics”, in the sense of the
philosophy of language, of the mathematization of physical contents, or of
the transformation of the objects in relations.33 If one considers the evolu-
tion of the question of the interpretation of quantum mechanics since the
first debates on the subject, and the subsequent familiarization acquired in
this area by physicists, up to the new knowledge made available during the
last years and to the reinterpretations that may be formulated consequently,
it is tempting to speak, on the contrary, of a semantization of the syntax.
The description of quantum physical systems was conceived previously only
by means of operations, while henceforth (by means of the transformations

31
It is possible in such a way to “think quantum non-separability” (Paty [63]).
32
Paty [85].
33
The expression is from Ernst Cassirer, cf. Cassirer [13]. See Granger [33].
60 M. PATY

alluded to above in our definitions of the involved physical magnitudes) it


might evolve toward an interpretation in terms of physical systems conceived
as objects possessing properties: this being at the expense of modifying our
definitions of what is a physical quantity for the description of such prop-
erties. The previous “syntax” remained exterior to the physical content it
concerned, insofar that it considered itself to be strictly confined to purely
formal means of the description, without declaring a definite position as to
the physical feature of this physical content, and even its very existence. A
full achievement of the program of the semantization of the theory would
consist of formulating from the start on the quantum theory as the theory
of a category of physical objects and of their physical properties. This would
be a necessary task before entering again upon new syntaxizations. (One
should be able to substitute to the axiomatic formulation à la von Neu-
mann 34 an equivalent axiomatic formulated in terms of physical properties
concerning physical entities at the quantum level).

6. ALGORITHM AND NATURALISTIC REDUCTION OF


REPRESENTATION

The notion of epistemic operation quite naturally leads to ask to what ex-
tent such operations can be expressed by algorithms. In the case of simple
operations, such as the search of invariants, it is perhaps possible to imagine
equivalent algorithms. However, though an algorithm for the construction of
an already known invariant might be found, this does not entail the possibil-
ity of also an algorithm that would have led to the discovery of that invariant.
For complex operations bearing on the acquisition of new and more precise
knowledge, the opinions concerning the possibility of algorithms diverge.
The experts in artificial intelligence and its adepts will willingly reply af-
firmatively. According to them, in the long range, any cognitive operation,
scientific inventions included, will be reconstituted by machines: for sim-
ple operations they are already proposing models, and they are proclaiming
an obvious necessity of principle in any case, sending skeptics back to the
archaic matter-mind dualism.35 Why, do they claim, should we hold that
brain is essentially different from a machine, a neuronal machine?
Of course, by certain aspects, this attitude has ancient precedents.
The Cartesian research of a method to get a certain knowledge, Leibniz’s
search of a modality for formulating the totality of knowledge in a universal
and perfectly logical language, and even the parti-pris, in Spinoza’s Ethics,
of a more geometrico proof of each statements, not to go back still farther
34
Neumann [57].
35
Changeux [16], Damasio [21].
EPISTEMIC OPERATIONS AND FORMALIZED EPISTEMOLOGY 61

until Aristotle, testify of the permanence of similar concerns the history of


philosophy; stemming from a belief in the possibility of a powerful algorithm
permitting to found, to grasp and to organize the totality of “true” knowl-
edge. This quest of synthesis by means of formal unity does not necessarily
possess a reductionist character: in Descartes’ view it was protected by the
dualism (acted matter)-(thinking mind), and the Spinozian monism left full
room for the specificities of all the different sciences. But in the case of the
modern conception of the brain as a machine, a character of a quite different
nature is involved: what is imagined in this case is not only an algorithm
for the representation of any knowledge, but furthermore also a naturalistic
reduction of the processs of generation of knowledge,36 a question which
cannot be discussed here.
The claim for algorithms of production of the scientific knowledge
has been strengthened in our time by the logicist views of the philosophers
from the Vienna and the Berlin circles 37 and of their successors, dissidents
or not, up to analytic philosophy.38 According to the logical positivists
and empiricists, science is doomed to generate a compelling philosophy of
knowledge (the so called “scientific philosophy”, exposed and supported by
Hans Reichenbach39 ), a philosophy rooted in the experience data taken as
the fundamental reference for knowledge. The search, with Rudolf Car-
nap, for an inductive logic, be it only a probabilistic one,40 presupposed
the idea that any scientific knowledge can be reduced to rules valid every-
where, perhaps in all times, which have to be discovered: which amounts
to assigning to scientific knowledge an algorithmic essence. The assertion of
confidence in methodology,41 and of the normative legitimacy of a “rational
reconstruction” of the scientific knowledge (Reichenbach, Popper,42 among
others) that would allow to collect the irrationalities due to the interven-
tion of the subject in a knowledge of which the vocation is to be objective,
so a “knowledge without subject”, are other attitudes that go in the same
direction.
The Popperian “third world” of the forms of objective knowledge 43
can also be regarded as an indication (or an effect) of this sort of view: such
36
See the quite interesting dialogue between Jean-Pierre Changeux and Paul Ricœur
(Changeux and Ricœur [17]).
37
Wiener Kreis [19], Hahn, Neurath and Carnap [36], Soulez [111].
38
See Joelle Proust’s book on “questions of form” of logic and of analytical statements,
from Kant to Carnap (Proust [103])
39
Reichenbach [105,106].
40
Carnap [10,11], Jeffrey [39].
41
Bunge [8]. Alberto Cupani (Cupani [20]) rightly reminds us, however, that, for Bunge,
the method is not a “recipe” that would be mechanically or automatically applied.
42
Reichenbach [104,105,106,107], Popper [99,100].
43
Popper [100], p. 154.
62 M. PATY

an impersonal universe of ideas is supposed to be that of pure rationality,


cleared of affect as well as of chance, and even of materiality (these being
sent back, respectively, to the second world and to the first one). This world
of created forms, without the acts of creation, is akin to a reservation or a
museum: a “museum of ideas”, “virtual”—before the recent common use
of this term—where one can draw prime matter from for formulating other
ideas, is conceivable in fact only as organically related to the two other
Popperian worlds. Karl Popper’s third world seems to harbour a rather
Platonian desire to purify the world of ideas from perishable elements such
as the matter, flesh, affects, feelings and aims of which the individual subject
is made, while also protecting a “logic of reconstruction” that is not very
different from an algorithmic function.
This being said, we shall however recognize that an algorithm devoted
to a logical application cannot be identified per ipse to a machine, because
it does not necessarily involve the condition of reproducing the totality of
the cognitive operations. Furthermore, one also should take into account a
certain widening of the concept of machine, to include the possibility of new,
emergent forms or properties concerning material systems as well as those
from the space of ideas. But invoking a powerful algorithm or a machine,
even with organic properties, as being able to reproduce or to describe the
process of acquisition of a fundamentally new representation, gives rise to
reserves, without any need to invoke some dualism, but on the contrary
by holding an ontologically monist position: these reserves are similar to
those which one can soundly oppose to reductionism and to a naturalistic
conception of knowledge and of values. It is possible to emit them without
for that denying any interest to the concept of epistemic operations.
It is possible to conceive an epistemic operation to be the source of an
algorithm, such as the invariants considered above, or—another example—
such as the Leibnitz differential calculus; and, once the algorithm has been
invented, to permit to reconstitute or to reorganize a representation of it
inside some chosen referential of meanings, with all the ascertained or pre-
dictable properties of such a representation. The algorithm could demon-
strate its fruitfulness in the resolution of many problems, and it even could
perhaps contribute to formulate new problems and to solve them. But can
it be also conceived to generate, by itself, an essentially, a qualitatively new
property or knowledge? This seems to be possible only if the algorithm con-
tained in itself this break relative to its antecedents, that makes the new.
But would we not, thereby, have already deserted the bounded framework
inside which epistemic operations can be formulated?
So far as we know, the machine for producing conceptual novelty still
remains to be invented. Does this entail that such a possibility is absolutely
EPISTEMIC OPERATIONS AND FORMALIZED EPISTEMOLOGY 63

unthinkable for the future? The answer to this question, which artificial
intelligence would like to be positive, depends on what can be said to be
“qualitatively new”: this characteristic certainly escapes the strictly inter-
nal content of a given knowledge, i.e., what in it can be formulated by ref-
erence to exclusively its own conceptual framework and inside the universe
of meanings in which it is immersed. It is difficult to imagine the existence,
or even only the possibility of a “machine for producing meaning”, in the
common acceptation of these terms, in the absence of a mind that would be
at the origin of this meaning or that could “read” it.
Evaluating concepts and their possible character of novelty is of the
order of signification, and as for now it is the human mind, tied with a brain
inseparable from a body and from a practice of life, and setting aims for it-
self, by will or by desires, which imposes its meanings upon the machines. A
“machine for produce sense” would have to possess the properties mentioned
above, and certainly also others, including psychology and feelings: such a
machine would astonishingly look like a man in society, whose generation by
nature is the result of a very long—and maybe improbable—history the ori-
gin of which is lost in the night of times: a history of maturation, renewals,
transmissions and exchanges, resulting from the diversification of the human
phylum, biological, social and cultural, and from the accidents brought forth
by chance. So the fundamental question is the following one: is it possible
to conceive of an algorithm able to generate, for knowledge, meanings that
would differ from all the already available ones and would appear to us to be
legitimate, and even perhaps more certain or important? With questions of
this type, we trigger, it seems, an unending chain of implications and an infi-
nite multiplicity of open ways which a machine would have much difficulty to
solve. While the human mind, issued from matter, does not calculate on all
the possibilities as a machine does, it cuts across the available combinations
and makes choices long before having exhausted them.44 It simply posits
the meaning which—according to its judgment that might be “subjective”
in only a restricted sense—brings forth for it the sudden illumination of an
intelligibility. This—be it Cartesian evidence, Spinozian knowledge of the
third kind, illumination of the intuition in the sense of Poincaré, Einstein,
and other contemporary thinkers—seems doomed to durably escape any re-
ductionist representation or conception. This is so because a definite piece
of intelligibility, in order to gain a ground, must call in other intelligibilities
from an endless regressive chain. Like in Pascal’s considerations on the situ-
ation of human intelligence in this world, that first lean on reasons which he
believes to clearly understand by his experience but then, when questioned,
44
As Poincaré noted when speaking of the “choice of the [significant] facts” (Poincaré
[94]).
64 M. PATY

appear to draw down into a bottomless well.45

REFERENCES
1. A. Ashtekhar, “Non-perturbative quantum gravity: A status report,”
in M. Cerdonio and R. Rianci, eds., General Relativity and Gravitation
(World Scientific, Singapore, 1989).
2. J. Audoze, P. Musset, and M. Paty, eds., Les particules et l’univers
(Presses Universitaires de France, Paris, 1990).
3. F. Bailly, “About the emergence of invariants in physics: From ‘sub-
stantial’ conservation to formal invariance,” in Mugur-Schächter and
van der Merwe [56].
4. F. Bailly, “Remarks about the program for a formal epistemology,” in
Mugur-Schächter and van der Merwe [56].
5. B. Barnes and D. Bloor, 1982, “Relativism, rationalism and the soci-
ology of knowledge,” in Hollis, Luckes, 1982, pp. 21-47.
6. J. S. Bell, Speakable and Unspeakable in Quantum Mechanics (Univer-
sity Press, Cambridge, 1987).
7. L. M. Brown, A. Pais, and B. Pippard, eds., Twentieth Century
Physics, Vols. 1,2,3 (Philadephia Institute of Physics, New York,
1995).
8. M. Bunge, Philosophy of Physics (Dordrecht, Reidel, 1973).
9. M. Bunge, Epistémologie (Maloine, Paris, 1983).
10. R. Carnap, Die Aufgabe der Wissenschaftslogik (Vienna, 1934), trans-
lation of Le problème de la logique de la science (Hermann, Paris,
1935).
11. R. Carnap, Logical Foundations of Probability (Chicago University
Press, Chicago, 1950).
12. R. Carnap, Philosophical Foundations of Physics (Basic Books, New
York, 1966).
13. E. Cassirer, Substanzbegriff und Funktionsbegriff (Bruno Cassirer,
Berlin, 1910); translation: Substance and Function, in Cassirer [14],
pp. 1-346.
14. E. Cassirer, Substance and Function and Einstein’s Theory of Relativ-
ity, English translation by W. C. Swabey and M. C. Swabey (Open
Court, Chicago, 1923; Dover, New York, 1953).
15. E. Cassirer, Philosophie der symbolischeen Formen, 3 volumes (Berlin,
1923, 1925, 1929); transl. by O. Hansen-Love and J. Lacoste, La
philosophie des formes symboliques, 3 volumes (Minuit, Paris, 1972).
16. J. P. Changeux, L’homme neuronal (Fayard, Paris, 1983.)
45
Pascal [62].
EPISTEMIC OPERATIONS AND FORMALIZED EPISTEMOLOGY 65

17. J. P. Changeux and P. Ricœur, La nature et la règle. Ce qui nous fait


penser (Odile Jacob, Paris, 1997).
18. R. Cohen, M. Horne, and J. Stachel, eds., Experimental Metaphysics:
Quantum Mechanical Studies for Abner Shimony, Vol. One (Kluwer
Academic, Dordrecht, 1997).
19. R. Cohen, M. Horne, and J. Stachel, eds., Potentiality, Entanglement
and Passion-at-a-Distance: Quantum Mechanical Studies for Abner
Shimony, Vol. Two (Kluwer Academic, Dordrecht, 1997).
20. A. Cupani, “Julgamento cientifico e racionalidade,” in Dutra, ed., Nos
limites da Epistimolegia Analitica (NEL, Florianopolis, 1999), pp. 139-
159.
21. A. R. Damasio, Descartes’ Error. Emotion, Reason, and the Human
Brain (Grosset, 1994).
22. P. A. M. Dirac, The Principles of Quantum Mechanics (Clarendon,
Oxford, 1930; 4th edn., 1958; revised 1981).
23. A. Einstein, “Spezielle Relativitätstheorie” (manuscript on the spe-
cial theory of relativity, unpublished because of the war, prepared in
1912 for the Handbuch der Radiologie, E. Marx, ed.; copy at Einstein
Archives), in Ref. [24], Vol. 4, pp. 3-108.
24. A. Einstein, The Collected Papers of Albert Einstein, Vols. 1-8, J.
Stachel et al., eds. (University Press, Princeton, NJ, 1987-1998).
25. A. Einstein, H. A. Lorentz, H. Minkowski, and H. Weyl, Das Rela-
tivitätsprinzip, 4th edn., A. Sommerfeld, ed. (Teubner, Leipzig, 1922);
translation by W. Perrett and G. B. Jeffery, The Principle of Relativity
(Methuen, London, 1923; Dover, New York, 1952).
26. A. Einstein, B. Podolsky, and N. Rosen, “Can quantum-mechanical
description of physical reality be considered complete?,” Phys. Rev.
47, 770-780 (1935).
27. B. d’Espagnat, Le réel voilé. Analyse des concepts quantiques (Fayard,
Paris, 1994).
28. H. Focillon, La Vie des Formes, suivi de Eloge de la main (Presses
Universitaires de France, Paris, 1981, 1990).
29. M. Foucault, L’archéologie du savoir (Gallimard, Paris, 1969).
30. R. Giere, Explaining Science. A Cognitive Approach (University Press,
Chicago, 1988).
31. G. G. Granger, Essai d’une philosophie du style (Armand Colin, Paris,
1968; Odile Jacob, Paris, 1988).
32. G. G. Granger, Pour la connaissance philosophique (Odile Jacob,
Paris, 1988).
33. G. G. Granger, Formes, opérations, objets (Vrin, Paris, 1994).
34. G. G. Granger, Le probable, le possible et le virtuel (Odile Jacob, Paris,
1995).
66 M. PATY

35. J. Hadamard, An Essay on the Psychology of Invention in the Math-


ematical Field (University Press, Princeton, 1945).
36. H. Hahn, O. Neurath, and R. Carnap, Wissenschaftlische Weltauflas-
sung der Wiener Kreis (Vienna, 1929); reprinted in Neurath [1979],
pp. 81-101; see [120].
37. W. R. Hamilton, “On a general method in dynamics,” Phil. Trans.
Roy. Soc. (London) 1934; reprinted in [38], Vol. 2.
38. W. R. Hamilton, The Mathematical Papers of Sir William Rowland
Hamilton, 3 volumes (University Press, Cambridge, 1931, 1940, 1967).
39. R. C. Jeffrey, Studies in Inductive Logic and Probabilities, 2 volumes
(Berkeley, 1980).
40. I. Kant, Kritik der reinen Vernunft (Hartknoch, Riga, 1781); 2nd mod-
ified edn., 1787; translation by Alexandre J. L. Delamarre and Fra-
nois Marty, “Critique de la raison pure,” in Emmanuel, Kant, Oeuvres
philosophiques, Vol. 1 (Gallimard, Paris, 1980), pp. 705-1470.
41. J. Kouneiher, Epistémologie et histoire récente de la gravitation quan-
tique, Doctoral Thesis in Epistemology and History of Sciences, Uni-
versité Paris 7-Denis Diderot, 1998.
42. J. L. Lagrange, Mécanique analytique (Desaint, Paris, 1788; 2nd edn.,
1811); reprinted in Oeuvres complètes de Lagrange, Vols. 11 and 12,
published by J. A. Serret and G. Darboux (Gauthier-Villars, Paris,
1889).
43. I. Lakatos, Philosophical Papers, Vol. 1: The Methodology of Scien-
tific Research Programmes, J. Worrall and G. Currie, eds. (University
Press, Cambridge, 1978).
44. I. Lakatos, and A. Musgrave, eds., Criticism and the Growth of Knowl-
edge: Proceedings, International Colloquium in the Philosophy of Sci-
ence, London, 1965, Vol. 4 (University Press, Cambridge, 1970).
45. A. Lalande, Vocabulaire technique et critique de la philosophie (1926),
13th edn. (Presses Universitaires de France, Paris, 1980).
46. P. Langevin, “L’évolution de l’espace et du temps,” Scientia (Bologna)
10, 31-54 (1911) (conférence au Congrès de philosophie de Bologne,
1911); reprinted in Langevin [48], pp. 265-300.
47. P. Langevin, “Le temps, l’espace et la causalité dans la physique mod-
erne,” Bulletin de la Société francaise de philosophie 12, 1-46 (1911);
reprinted in Langevin [48], pp. 301-344.
48. P. Langevin, La physique depuis vingt ans (Doin, Paris, 1923).
49. C. Levi-Strauss, Anthropologie structurale (Plon, Paris, 1958).
50. H. Margenau, Physics and Philosophy: Selected Essays (Reidel, Dor-
drecht, 1978).
51. H. Minkowski, “Das Relativitätsprinzip,” lecture in Göttingen,
5.11.1907; posthumous publication arranged by A. Sommerfeld, Ann.
Phys. 47, 927-938 (1915).
EPISTEMIC OPERATIONS AND FORMALIZED EPISTEMOLOGY 67

52. H. Minkowski, “Raum und Zeit,” lecture at the 80th Meeting of Ger-
man scientists and physicians, Cologne 21.9.1908; Physikal. Z. 10, 104-
111 (1909). “Space and Time,” in Einstein et al. [25], pp. 73-91.
53. M. Mugur-Schächter, “From quantum mechanics to universal struc-
tures of conceptualization and feedback on quantum mechanics,”
Found. Phys. 23, 37-122 (1993).
54. M. Mugur-Schächter, “Une méthode de conceptualisation relativisée:
vers une épistémologie formelle apte à faire face aux complexités,”
Revue Internationale de Systémique 9 (2), 269-303 (1995).
55. M. Mugur-Schächter, “Les leçons de la mécanique quantique (vers une
épistémologie formelle),” Le Débat 94, 169-192 (1996).
56. M. Mugur-Schächter and A. van der Merwe, eds., this book (Kluwer
Academic, Dordrecht, 2003).
57. J. von Neumann, Mathematische Grundlagen der Quantenmechanik
(Springer, Berlin, 1932); transltion by Alexandre Proca, Les fonde-
ments mathématiques de la mécanique quantique (Librairie Alcan &
Presses Universitaires de France, Paris, 1947).
58. E. Nœther, “Invarianten beliebiger Differentialausdrücke,” Nachr.
Königl. Ges. Wiss. Göttingen, Math. Phys. Kl., 1918, pp. 235-257;
in Noether [60], pp. 240-247.
59. E. Nœther, “Invariante variationprobleme,” ibid., pp. 235-257;
reprinted in Noether [60], pp. 248-270; translation by M. A. Tavel,
Transport Theory and Statistical Physics 1, 183-207 (1971).
60. E. Nœther, Collected Papers, N. Jacobson, ed. (Springer, Berlin, 1983).
61. OULIPO (Ouvroir de Littérature potentielle), Littérature potentielle:
créations, re-créations, récréations (Gallimard, Paris, 1977).
62. B. Pascal, “De l’esprit géométrique,” in Pascal, Oeuvres complètes,
Présentation et Notes de Louis Lafuma (Seuil, Paris, 1993), pp. 348-
355.
63. M. Paty, “La non-séparabilité locale et l’objet de la théorie physique,”
Fundamenta Scientiae 7, 47-87 (1986).
64. M. Paty, La matière dérobée. L’appropriation critique de l’objet de la
physique contemporaine (Archives contemporaines, Paris, 1988).
65. M. Paty, “Sur la notion de complétude d’une théorie physique,” in N.
Fleury, Joffily, Sergio, J. A. Martins Simes, and A. Troper, eds., Leite
Lopes Festchrift: A Pioneer Physicist in the Third World; dedicated
to J. Leite Lopes on the occasion of his seventieth birthday (World
Scientific, Singapore, 1988), pp. 143-164.
66. M. Paty, L’analyse critique des sciences, ou le tétraèdre épistémologique
(sciences, philosophie, épistémologie, histoire des sciences) (L’Harmattan,
Paris, 1990).
68 M. PATY

67. M. Paty, “L’endoréférence d’une science formalisée de la nature,” in


C. Dilworth, ed., Intelligibility in Science (Rodopi, Amsterdam, 1992),
pp. 73-110.
68. M. Paty, Einstein philosophe. La physique comme pratique philosophique
(Presses Universitaires de France, Paris, 1993).
69. M. Paty, “Mesure, expérience et objet théorique en physique,” in J.-C.
Beaune, ed., La mesure, instruments et philosophies (Champ Vallon,
Seyssel, 1994), pp. 159-174.
70. M. Paty, “Le caractère historique de l’adéquation des mathématiques
à la physique,” in S. Garma, D. Flament, and V. Navarro, eds., Con-
tra los titanes de la rutina—Contre les titans de la routine (encuen-
tro en Madrid de investigadores hispano-franceses sobre la historia
y la filosofia de la matematica Rencontre à Madrid de chercheurs
hispano-franais sur l’histoire et la philosophie de la mathématique,18-
22.11.1991) (Comunidad de Madrid/C.S.I.C., Madrid, 1994), pp. 401-
428.
71. M. Paty, “Sur l’histoire du problème du temps: le temps physique
et les phénomènes,” in E. Klein and M. Spiro, eds., Le temps et sa
flèche (Frontières, Gif-sur-Yvette, 1994; 2nd edn., 1995). (Collection
Champs, Flammarion, Paris, 1996), pp. 21-58.
72. M. Paty, “The nature of Einstein’s objections to the Copenhagen in-
terpretation of quantum mechanics,” Found. Phys. 25, 183-204 (1995).
73. M. Paty, “Poincaré et le principe de relativité,” in J.-L. Greffe,
G. Heinzmann, and K. Lorenz, eds., Henri Poincaré. Science et
philosophie. Science and philosophy. Wissenscchaft und Philosophie.
Congrès international, Nancy, France, 14-18 mai 1984 (Akademie-
Verlag, Berlin/Albert Blanchard, Paris, 1996), pp. 101-143.
74. M. Paty, “Le style d’Einstein, la nature du travail scientifique et le
problème de la découverte,” Revue philosophique de Louvain 94, 447-
470 (1996).
75. M. Paty, “Predicate of existence and predictivity for a theoretical
object in physics,” in E. Agazzi, ed., Realism and Quantum Physics
(Rodopi, Amsterdam, 1997), pp. 97-130.
76. M. Paty, “Philosophie et physique,” in J.-F. Mattéi, ed., Le Discours
philosophique, volume 4 de l’Encyclopédie philosophique universelle
(Presses Universitaires de France, Paris, 1998), pp. 2104-2122.
77. M. Paty, “Universality of science: Historical validation of a philosoph-
ical idea,” in S. I. Habib and D. Raina, eds., Situating the History
of Science: Dialogues with Joseph Needham (Oxford University Press,
New Delhi, 1998), Chap. 12.
78. M. Paty, “La place des principes dans la physique mathématique au
sens de Poincaré,” in J. Sebestik and A. Soulez, eds., “Actes du Col-
EPISTEMIC OPERATIONS AND FORMALIZED EPISTEMOLOGY 69

loque France-Autriche Paris, mai 1995, Interférences et transforma-


tions dans la philosophie francaise et autrichienne (Mach, Poincaré,
Duhem, Boltzmann),” Fundamenta philosophiae (Nancy) 3 (2), 61-74
(1998-1999).
79. M. Paty, “La création scientifique selon Poincaré et Einstein,” in
M. Serfati, ed., La recherche de la vérité (Collection Ecriture des
Mathématiques, Éditions ACL, Paris, 1999), pp. 241-280.
80. M. Paty, “Paul Langevin (1872-1946), la relativité et les quanta,” Bul-
letin de la Société Franaise de Physique 119, 15-20 (1999).
81. M. Paty, “Are quantum systems physical objects with physical prop-
erties?,” Eur. J. Phys. 20, 373-388 (1999).
82. M. Paty, “Cosmologie et matière quantiquea,” in J. Seidengart and
J.-J. Sczeciniarz, eds., Cosmologie et philosophie. En hommage à
Jacques Merleau-Ponty; Epistémologiques (Paris, Sao Paulo) 1, 219-
249 (2000).
83. M. Paty, “Interprétations et significations en physique quantique,” Re-
vue Internationale de philosophie 54, 199-242 (2000).
84. M. Paty, “Langevin, Poincaré et Einstein,” in Paul Langevin, son œu-
vre et sa pensée. (Paris, São Paulo, forthcoming).
85. M. Paty, “The quantum and the classical domains as (provisional)
parallel coexistents,” in S. French, D. Krause, and F. Doria, eds.
(festschrift in honour of Newton da Costa, on the occasion of his sev-
entieth birthday), Synthese 125, 179-200 (2000).
86. M. Paty, “The concept of quantum state: New views on old phenom-
ena,” in R. S. Cohen, and J. Renn, eds., John Stachel Festschrift
(Boston Studies in the Philosophy and History of Science, Kluwer,
Dordrecht, forthcoming).
87. M. Paty, “The idea of quantity at the origin of the legitimacy of math-
ematization in physics,” in C. Gould, ed., Conference on the Philoso-
phy of Marx Wartofsky, New School University, New York, March 6,
1999.
88. M. Paty, “L’analogie mathématique au sens de Poincaré et sa fonction
en physique,” in M.-J. Durand-Richard, ed., Le statut de l’analogie
dans la démarche scientifique (Editions du CNRS, Paris, forthcoming).
89. M. Paty, Einstein, les quanta et le réel. Critique et construction
théorique, in the press.
90. G. Perec, La disparition (Denoël, Paris, 1969; Gallimard, Paris, 1989).
91. H. Poincaré, “Sur les rapports de l’analyse pure et de la physique
mathématique,” Acta Mathematica 21, 331-341 (1897); reprinted in
Poincaré [98], pp. 17-30; “Les rapports de l’analyse et de la physique
mathématique,” Revue générale des sciences pures et appliquées 8,
70 M. PATY

857-861 (1987); reprinted in Poincaré [93], Chap. 5 (“L’analyse et la


physique”), 1970 edn., pp. 103-113.
92. H. Poincaré, “Sur la dynamique de l’électron” (received July 23, 1905),
Rendiconti del Circolo matematico di Palermo XXI 129-176 (1906);
also in Poincaré [97], Vol. 9, pp. 494-550.
93. H. Poincaré, Science et méthode (Flammarion, Paris, 1908, 1918).
94. H. Poincaré, “Le choix des faits,” The Monist, 1909, 231-232; reprinted
in Poincaré [93], Book 1, Chap. 1, 1918 edn., pp. 16-18.
95. H. Poincaré, “L’espace et le temps,” Scientia XXV, 159-170 (1912);
also in Poincaré [96], 1963 edn., pp. 97-109.
96. H. Poincaré, Dernières pensées (Flammarion, Paris, 1913, 1963).
97. H. Poincaré, Oeuvres, 11 volumes (Gauthier-Villars, Paris, 1950-1965).
98. H. Poincaré, L’analyse et la recherche, selection of texts with an in-
troduction by G. Ramunni (Hermann, Paris, 1991).
99. K. Popper, Logik der Forschung. Zur Erkenntnisstheorie der modernen
Naturwissenschaft (Springer, Wien, 1934); re-edited with additions,
1966; translation: The Logic of Scientific Discovery (Hutchinson, Lon-
don, 1959, 1968).
100. K. Popper, Objective Knowledge: An Evolutionary Approach (Clare-
don, Oxford, 1972).
101. K. Popper, Quantum Theory and the Schism in Physics, from the
Postscript to the Logic of Scientific Discovery, Vol. 3 (Hutchinson,
London, 1982; written in 1956).
102. K. Popper, Realism and the Aim of Science, from the Postscript to the
Logic of Scientific Discovery, Vol. 1, W. W. Bartley III, ed. (Hutchin-
son, London, 1983; written in 1956).
103. J. Proust, Questions de forme. Logique et proposition analytique de
Kant à Carnap (Fayard, Paris, 1986).
104. H. Reichenbach, Experience and Prediction (University Press, Chicago,
1938).
105. H. Reichenbach, The Rise of Scientific Philosophy (University of Cal-
ifornia Press, Berkeley, 1951; re-edited 1973).
106. H. Reichenbach, Modern Philosophy of Science, edited and translated
by M. Reichenbach (Routledge & Kegan Paul, London, 1959).
107. H. Reichenbach, Selected Writings, 2 volumes, R. Cohen and M. Re-
ichenbach, eds. (Reidel, Dordrecht, 1978).
108. D. Ruelle, Chaotic Evolution and Strange Attractors (University Press,
Cambridge, 1989).
109. D. Ruelle, Hasard et Chaos; Chance and Chaos (Uniersity Press,
Princeton, 1991).
110. A. Shimony, Search for a Naturalistic World View, 2 volumes (Univer-
sity Press, Cambridge, 1993).
EPISTEMIC OPERATIONS AND FORMALIZED EPISTEMOLOGY 71

111. A. Soulez, ed., Manifeste du Cercle de Vienne et autres écrits: Carnap,


Hahn, Neurath, Schlick, Waissmann, Wittgenstein (Presses Universi-
taires de France, Paris, 1986).
112. W. Stegmüller, Probleme und Resulate der Wissenschaftstheorie und
Analytischen Philosophie. Band 2: Theorie und Erfahrung. Zweiter
Halbband: Theoriestrukturen und Theoriendynamik (Springer, Heidel-
berg, 1973).
113. W. Stegmüller, The Structuralist View of Theories. A Possible Ana-
logue of the Bourbaki Programme in Physical Science (Springer, Hei-
delberg, 1979).
114. P. Veyne, Comment on écrit l’histoire (Seuil, Paris, 1971).
115. P. Veyne, “Foucault révolutionne l’histoire,” in Comment on écrit
l’histoire et Foucault révolutionne l’histoire (Seuil, Paris, 1978).
116. S. Walter, Hermann Minkowski et la mathématisation de la théorie de
la relativité, 1905-1915, Doctoral Thesis in Epistemology and History
of Sciences, Université Paris 7-Denis Diderot, 1996.
117. M. Wartofsky, Conceptual Foundations of Scientific Thought: An In-
troduction to the Philosophy of Science (Mac Millan, London, 1968).
118. H. Weyl, Gruppentheorie und Quantenmechanik (Hirzel, Leipzig, 1928,
2nd edn., 1931); translation by H. P. Robertson, The Theory of Groups
and Quantum Mechanics (Methuen, London, 1931; Dover, New York,
1950).
119. J. A. Wheeler, and W. H. Zurek, eds., Quantum Theory of Measure-
ment (University Press, Princeton, 1983).
120. Wiener Kreis, Wissenschaftliche Weltaufflassung: der Wiener Kreis
(Wolf, Wien, 1929).
121. E. Zahar, Einstein’s Revolution. A Study in Heuristic (Open Court,
La Salle, IL, 1989).
122. E. Zahar, “Againts the (alleged) divorce between science and philoso-
phy: Comments on M. Paty’s “La matière dérobée,” Rivista di Storia
della Scienza, Ser. II, 2, 191-208 (1994).
72 M. PATY
5
MATHEMATICAL PHYSICS AND FORMALIZED
EPISTEMOLOGY: DEBATE WITH JEAN PETITOT∗

Jean Petitot

Ecole Polytechnique
1 rue Descartes
75005 Paris, France

Interlocutors: Francis Bailly, Michel Bitbol,


Mioara Mugur-Schächter, Vincent Schächter

This paper develops an updated transcendentalist perspective concerning


the epistemological status of objectivity. The main point is that objectiv-
ity is neither an ontology nor a mere description of a phenomenal state of
affairs. It is instead a principled (categorial) “legalization” and a mathemati-
cal reconstruction of the phenomena. As the very concept of a phenomenon
is relational (relative to a receptive stance such as perception in classical
mechanics or measure devices in quantum mechanics), the conditions for
accessing them must be included in the very concept of objectivity. This
paper emphasizes the transcendental content of symmetries in modern phys-
ical theories, from general relativity to gauge invariance. A great deal of the
discussion focuses on epistemological key points in quantum mechanics.

Key words: physical objectivity, transcendental philosophy, mathematical


models, symmetries, quantum mechanics.

Mugur-Schächter. Jean Petitot, in order to present us a first insight


into your conception of “a formalized epistemology,” you have accepted to
summarize your views on mathematical physics.
Petitot. To begin with, I would like to stress that for me physics is
a factum rationis. Mathematical physics is a scientific fact, and in spite of
all its internal problems one has to accept that, in a sense that has to be
precisely defined, it yields an objective knowledge concerning a very large set
of phenomena. Physics works and requires therefore what I call a “plausible”
operational epistemology, that is, an epistemology where physical effectivity

See “Important Note” on p. xviii.

73
74 J. PETITOT ET AL.

occupies the master-position. In this respect, I separate from most of my


philosophers colleagues. I consider that philosophy does not have to rule
the status of objectivity. So, if physics is taken as an object of philosophical
investigations, it must occupy a leading position, and epistemology has to
take it as a datum. From this factum rationis, and without any pre-judgment
concerning the status of physical objectivity, it has to go on to problems of
another kind, ontological, cognitive, etc.
I would like to stress also right away a second point, that is quite
crucial: my theses are transcendentally oriented, in Kant’s sense. This in-
volves the key claim that physical objectivity is not an ontology: physics
doesn’t predicate on an “independent” reality but concerns exclusively phe-
nomena. Now, phenomena are by definition relational entities that exist
only with respect to a receptive device: no matter whether it is a measure
instrument or a perceptive living organism, a receptive device must exist.
This essential relativity of any phenomenon to a receptive instance has to
be included in the principles of physics. In quantum mechanics (QM) for
instance, a physical entity can be accessed only through convenient measur-
ing apparatus. In classical mechanics the role of the instruments is hold by
the perceptive system of the human observer (but this latter case is not so
different from the former one, since the visual system, and especially the
retina, is an extraordinary quantum instrument).
More generally, to uphold transcendental theses on physics amounts
simply to require the conditions for accessing physical phenomena—space,
time, instruments, etc.—to be included in the very concept of physical ob-
jectivity. I think that in the Bohr-Einstein controversy concerning this key
point, Bohr was fully right.
So physics deals exclusively with phenomena but possesses also an
undeniable objective content. But if physical objectivity is neither a subjec-
tive phenomenism nor a realist ontology, it must be something else.
The classical transcendental answer is that objectivity is a “legal-
ization” ruling phenomena. This thesis persists from Kant on, up to the
Vienna Circle. In fact, even if it was dogmatically anti-Kantian, the Vienna
Circle was transcendentally oriented and developed what has been called a
“grammatical” transcendentalism. Only the status of geometry in physics
(the celebrated debate on “synthetic a priori” geometrical frames) launched
a dramatic controversy against the Kantian tradition. But there was general
consensus on the fact:

(i) That physics deals only with phenomena and must include the rela-
tional conditions for accessing them.
(ii) That objectivity stems from a legalization of phenomena and that a
prescriptive and normative dimension is constitutive of objectivity.
MATHEMATICAL PHYSICS AND FORMALIZED EPISTEMOLOGY 75

Modern physicists have spontaneously rediscovered one of the main tran-


scendental thesis, namely that there exists a specific legality of the phe-
nomena as such. It is impossible to construct an objective explanation of
physical phenomena by starting with an absolute ontology of an “an sich”
independent reality. Objectivity can only be defined as an order of legality,
and such a prescriptive definition distinguishes essentially objectivity from
any ontology.
You see, objects of scientific knowledge are not given directly as such
with their phenomenon. It is why they have to be constituted by their
legalization, why, apart from its experimental descriptive dimension, any
scientific knowledge presupposes in its very principle also a prescriptive (nor-
mative) dimension.
Therefore, besides the metaphysical “ontological” difference between
phenomena and noumenal being (see Heidegger), there also exists the phys-
ical “objective” difference between phenomena and “object” of scientific
knowledge. In contradistinction to phenomena, “objects” exist only if quali-
fied according to norms, to eidetico-constitutive rules defining what Husserl
called a “regional objective essence”. The normative concept of an object
is a condition of possibility of any scientific activity. It anticipates and pre-
determines what belongs generally and typically to the phenomena of the
considered region.
The recurrent error of ontological realisms is to confuse the prescrip-
tive dimension of objectivity with an underlying “an sich” ontological reality,
which would at the same time exist in itself “behind” the phenomena, and
be experimentally inaccessible, theoretically unknowable, and, in spite of all
this, causally efficient.
If you agree with these preliminaries then the main question becomes:
how can the legalization be performed?
In Kant’s works concerning classical physics, the method consists es-
sentially in interpreting the general “categories” of objectivity by starting
from the “forms of givenness” of the phenomena, what he called the “pure
forms” of phenomenal manifestation. But, since an interpretation of the
categories of objectivity can be operational only if it is mathematical, these
pure forms themselves must also be mathematized. So, in any objectivity
conceived in the transcendental sense operates a mathematical hermeneutic
of the objective categories – what Kant called a “schematization” and, more-
over, a “construction” –, which eliminates the meaning of these categories
(their metaphysical use): this is the mathematical aspect of the obligation to
restrict the categories to empirical observables (what Kant called “transcen-
dental deduction”).1 One is thus naturally led to a grammatical apriorism
1
Here, the term “deduction” is not used in its logical sense but in its juridical sense of
76 J. PETITOT ET AL.

according to which mathematics provide the pool of syntaxes for legalizing


scientific objects. Moreover, those mathematics that do mathematize the
forms of manifestation determine also the appropriate type of syntax. In
short, transcendental philosophy is pre-eminently a philosophy of mathe-
matical physics. As was emphasized by Jules Vuillemin, it can be defined as
the philosophy taking into account the fact that the essence of objectivity
is mathematical.
Of course, I go very much further than Kant concerning legalization. I
hold that the ideal of mathematical science is, through the mathematization
of the forms of phenomenal givenness and the correlative “schematization-
construction” of the categories, to reconstruct the phenomena in principled
mathematical terms, to reproduce, to simulate them as exactly as possible.
This is conceivable precisely because phenomena are only phenomena. It
would be absurd to claim to reconstruct an ontologically independent reality
“in itself”. But if what physics explains is strictly phenomenal, then it is
fully legitimate to want to reconstruct what is given initially as a phenomenal
datum, and to substitute for it by a mathematical constructed structure.
And my thesis is that scientific explanation is precisely this. As Valéry said,
we know only what we are able to produce. If, for instance, using Newton’s
law, I reconstruct in a mathematical principled way the trajectories of all
the material bodies (from the falling bodies to the observed motion of the
planets), ipso facto I understand them. But this objective understanding
shares no ontological content. Its unique truth criterium is the accuracy of
the simulation.
Bailly. Then what difference do you make between reproduction via
Ptolemaic epicycles, and reproduction via the Kepler-Newton equations?
Petitot. Your question brings into play the core relation between al-
gorithms of reconstruction and the categories and principles of objectivity.
This relationship is the essence of the problem. Without adequate algo-
rithms it is impossible to reconstruct the endless diversity of phenomena.
And these algorithms, no matter of what could be their mathematical na-
ture, must be precise and endowed with a strong internal generativity. But,
on the other hand, in order to be explanatory, they must also be strongly
connected to the objective categories. There is therefore a fundamental dif-
ference between algorithms reconstructing phenomena by representing them,
and algorithms reconstructing phenomena by deducing them from law-like
principles. The difference between the reconstruction of the motion of the
planets by Ptolemaic epicycles and by Newton’s equation is a very good
example. In fact the epicycles algorithm is truly excellent. I have been told
that spatial agencies make still use of it. Like Fourier series, it is highly

justification (of the restriction of the application of the objective categories to phenomena).
MATHEMATICAL PHYSICS AND FORMALIZED EPISTEMOLOGY 77

performant. But it is only a mere representation of trajectories, an approx-


imation scheme. Its connection to the categories is missing and it lacks
therefore any explanatory power. For each case it requires a more or less ad
hoc analysis.
This is related to the problem of algorithmic complexity. The more
an algorithm is generative and “simple”, the more it is connected to basic
categories and principles, the more it is a “good” efficient one. In this
sense Newton’s law represents a fantastic progress. When one follows up
this transcendental line of thought it appears that epistemology of physics
must essentially focus on the form of the main equations which connect
deductively the categorial basis with the diversity of phenomena. If the
equations of the physical theories are not taken into consideration directly
and explicitly, then the whole transcendental conception breaks down and
one must go back to a more classical conception of epistemology, such as
it is nowadays generally practiced by most philosophers: one asks a lot of
questions on ontology, experience, observation, induction, truth, judgement,
etc., thus just applying general philosophy to physics, instead of working out
a plausible epistemology specifically adapted to what physics really is. To
place physics in a master position is to change our pre-conceived metaphysics
in order to bring it to an inner agreement with physics, not the converse.
Mugur-Schächter. Do you consider that at this point the main
contours of your view have been already expressed sufficiently for permitting
a relevant first discussion?
Petitot. I would like to illustrate my views with some examples
concerning both classical mechanics and quantum microphysics. But if you
prefer we can already debate on some points.
Mugur-Schächter. All right. Then I would like to make a remark
on “ontology”. I thoroughly agree with your exclusion from physics of any
attempt at the construction of an ‘independent’ ontology. This, in my view
also, is simply a self-contradicting expression since ‘ontology’ means ‘model
of the things’, any model involves qualifications, and ‘independent’ reality
such-as-it-is-in-itself is strictly unqualifyable by definition.
However I would not exclude any ontology. I hold that each physical
theory – besides the other sorts of only algorithmic mathematical models –
should also involve a ‘relative methodological ontology’ built in order to insure
intelligibility (relative: including the ineluctable relativities of phenomena,
to the receptive entities ; methodological: induced by the aim to achieve
something; ontology: a model concerning also ‘things’, as opposed to chains
of exclusively phenomena, or exclusively mathematical algorithms). Indeed,
if such a relative methodological ontology is absent at the explicit level,
as in nowadays quantum mechanics, the search for it never ceases before a
78 J. PETITOT ET AL.

satisfactory one is obtained.


A relative methodological ontology might come out to be very near
to the connection required by Kant between the mathematized forms of
‘donation’ of the phenomenal manifestations, and the categories of objec-
tivity. However, strictly speaking, it cannot be identified to this connection
because modern physics contains exceptions with respect to the Kantian pre-
supposition of a ‘donation’ of the phenomenal manifestations: in quantum
mechanics, for instance, such a ‘donation’ does not exist, in general.
Now, relative methodological ontologies, in the specified sense, are
not opposable to objectivity, quite the contrary. Often they are the basic
place where invariances are first required.
Petitot. The term “ontology” has such a wide range of uses that one
has to specify its meaning in each context. Personally, when I say “ontol-
ogy” I have in mind not models of phenomena but a supposedly adequate
conceptual description of an independent substantial reality. I think with
Kant that “ontology” is a scientifically ill-formed concept (a category mis-
take). My rejection of ontology parallels that made by quantum physicists
such Bernard d’Espagnat when they stress that quantum objectivity is only
a “weak” objectivity, and cannot be, for principled reasons, a “strong” ob-
jectivity predicating about an independent reality in itself.
Now the question is: if we reject ontology are we necessarily com-
mitted to any sort of subjective idealism? If objectivity is divorced from
substantial ontology and constitutively depends upon the forms of phenom-
enal givenness and accessibility, does this brings necessarily back to the
subject and his cognitive processes? Not at all indeed. Let us take the
case of QM. By construction, quantum objects can exist only in so far as
they are measured and experimentally accessed. So the necessity of mea-
suring apparatus must be included in the very concept of quantum object.
However this does’nt entail that any specific theory of any specific sort of
apparatus must be part of the axioms of QM. The relations to instruments
making observable the phenomena is a sort of generalized principle of rel-
ativity. For instance, Heisenberg’s principle is totally independent of any
specific measuring device. It takes into account the unavoidable necessity of
a “receptive” device, but at the same time it brackets all the peculiarities
of such a device.
It is the same thing with classical mechanics. There, the restriction
to observables means that mechanics can speak only of space-time trajecto-
ries, and not of any metaphysical “substance” that would exist outside space
and time. One must not forget that at that time by ontology was meant
for instance a Monadology in the sense of Leibniz. Philosophers thought
that space and time were imaginary subjective entities existing only in our
MATHEMATICAL PHYSICS AND FORMALIZED EPISTEMOLOGY 79

minds. Ontology was foreign to space and time, and so physics could’nt be
“objective” in a strong sense. Kant’s answer to this antinomy was to look at
the way geometry operates in physics. Geometry concerns space as a subjec-
tive form. But at the same time there is absolutely no psychological element
in it. Neurobiology, psychophysics, psychology are not part of the axioms
of geometry. We can say that mathematical geometry is “ideal”, completely
disembodied. Space and time are “pure” forms. I say that they are universal
formats for brain information processings. Once you have mathematized
them, you can say that physical phenomena are spatio-temporal entities re-
lated with geometry. Then geometry prescribes constraints to objectivity,
e.g., principles of relativity acting on positions, velocities, and motions.
To summarize, on the one hand, an objective theory has to incor-
porate the fact that phenomena presuppose a receptive instance, but on
the other hand it must also bracket any particular theory of this receptive
instance.
Mugur-Schächter. It is not canceled. It is maintained, but – ex-
clusively – in its full generality with respect to the whole diversity of the
receptive entities utilized by human investigators in their various investigat-
ing actions of the considered domain of phenomena. It is maintained in so
far that it is ‘objectively subjective’, with respect to human investigators.
This is a relative zero point, not an absolute one.
Petitot. Exactly, I agree.
Mugur-Schächter. So you don’t cancel it.
Petitot. All right. It remains as a trace. For instance, general
relativity reduces the “subject” to an “observer”. But what is an “observer”?
Just a mere reference frame.
Bailly. It is inter-subjectivity.
Petitot. It is a relativity group. “Subjectivity” has become a group
of invariance. It is completely disembodied.
Mugur-Schächter. Not only one, indefinitely many. The ‘observer’
is the sum of all such known or potential human groups of relativity. (The
‘conceptor’ is still much more than this, but this is another topic). So the
human ‘subject’ is not in the least canceled from physics. In particular
he is not canceled from neither relativity nor quantum mechanics. He is
there, and not as a ‘trace’, not as a sign of something that was there but
ceased to be, he is there actually, as a quite positive, active, rich, dense,
specific general extract. And it is noteworthy that the extract would be
different in a science built by another sort of living beings. Another sort of
being might experience other forms of phenomenal donation and also impose
other categories of objectivity.
Petitot. All right, but I think that we must go still much further
80 J. PETITOT ET AL.

in this direction. I think that the role of relativity groups and symmetries
raises a very deep philosophical question. As was emphasized by Daniel
Bennequin, mathematical physics is “Galoisian” in the sense that what you
can know is negatively determined by what you can’t know. Symmetries
express non-physical entities (an absolute origin of space or time, an abso-
lute direction, etc.). They group such indiscernable entities in equivalence
classes. But as far as they are the key of the mathematical reconstruction
of the phenomena they are priors determining the positive contents of the
theory. So the positive contents, what you can know, are deductive conse-
quences of what you can’t know. This is one of the most beautiful features
of mathematical physics.
Bailly. What can be known does not belong to the same domain as
that what cannot be known. What cannot be known bears on changes of
the referentials and frameworks, while what can be known concerns what
happens inside the frameworks.
Mugur-Schächter. What can be known concerning the studied
system is modeled so as to stay consistent with the properties assigned by
construction to the referential. But these last properties are defined so that
the description of the studied system shall become as general and synthetic
as possible. The referentials are very free conceptual constructs, not facts.
The mathematical representations of directly perceivable phenomena, also,
can be removed very far away from these perceivable phenomena. Thus
there develops a to and fro process of representation, re-representation, etc.,
of the observable phenomena, in the course of which the referentials and the
initial mathematical representation of the forms of phenomenal ‘donation’,
are both radically changed, under constraints of global logico-mathematical
coherence.
Petitot. Yes, but to grasp the point let us examine an example. In
quantum field theory, Feynman’s integrals provide an incredibly powerful
algorithm. But what is their formal genesis? One starts from symmetries
to construct (quite univoquely) a Lagrangian. According to the axioms of
QM this Lagrangian leads to a path integral. Then using various algorithms
(perturbative expansions, Wick’s theorem, the stationary phase principle,
the renormalisation group, etc.), one can construct models which can be
confirmed up to many decimals! There, one really attend the mathematical
genesis of physical contents from prior symmetries.
Mugur-Schächter. That is true. But the concept of a Lagrangian,
and the Feynmann integral itself, contain already essences drawn from
phenomena, via classical mechanics, elecromagnetism, quantum mechanics.
How were all these prerequisites established? As far as I know there does
not yet exist a worked out general method ruling the sort of operations
MATHEMATICAL PHYSICS AND FORMALIZED EPISTEMOLOGY 81

that lead to the stage when it becomes possible to take maximal advantage
of the symmetries which the nature of the involved referential and of the
defined mathematical descriptors imposes upon a final mathematical repre-
sentations of phenomena. To elaborate such a method could be a major aim
for a formalized epistemology.
Petitot. I agree. But I think that the extraordinary efficiency of the
constraints imposed upon physical phenomena by mathematical formalisms
has not yet been given enough importance. Nor this idea that what can
be known must be generated by what cannot be known, and is therefore
negatively determined.
Mugur-Schächter. A rather Popperian sort of idea.
Schächter. The approach which you describe can be perceived both
as negative or as positive: you start with some mathematical reconstruc-
tion of the forms of donation of the phenomenon, which is positive. Then
you generalize maximally this first mathematical reconstruction by impos-
ing all the symmetries required by the utilized referential, which amounts
to introducing the ‘negative’ knowledge afterward, as a sort of correction or
improvement. In this way, out of the initial ‘positive’ core of representation,
one elaborates the final mathematical representation of the phenomena as
the maximal class of equivalence constructible from the initial ‘positive’ rep-
resentation, which again is ‘positive’. This, moreover, is probably the order
in which the representations do indeed evolve most frequently. A posteriori
it is possible of course to reconstruct in the converse order also.
Petitot. Yes, but what I want to stress here is the very singular role
of mathematics. The procedures of mathematical physics are radically differ-
ent from those of common sense. Contrary to common wisdom, namely that
there exists a continuity between common sense and mathematical physics,
I think that the conceptual abstractions to which we submit the phenomena
in current life are not at all of the same type as those from mathematical
physics.
Bitbol. However a well known maxim asserts that any determination
is “negative”.
Mugur-Schächter. Expanding on the important remark of Francis
Bailly that ‘what can be known does not belong to the same domain as
what cannot be know’, and of Vincent Schächter’s remark on ‘negative’ and
‘positive’ knowledge, I would like to submit a conjecture.
The formulations in terms of ‘negative’ and ‘positive’ knowledge have
a paradoxical flavor, whereby they catch and mobilize the attention, which
no doubt is useful. But I suspect that a thorough analysis might reveal at
the bottom a matter of purely logico-mathematical self-consistency of the
global representation.
82 J. PETITOT ET AL.

Consider the simplest example, of position in space. One can a priori


imagine this specification in only two sorts of contexts, either in connection
with a space referential, or in connection with absolute space (in the Kantian
sense of a form of pure intuition), tertium non datur. Now, in absolute space,
the notion of ‘absolute position’ simply does not exist, it is not constructible,
it contradicts the concept of absolute space because there – by definition –
no origin is specified. Given that in the posited context an absolute position
is a logically impossible concept, to say that in absolute space one cannot
‘know’ or ‘distinguish’ an ‘absolute position’ is somewhat misleading.
Whereas a relative mutual qualification of two positions is possible in
absolute space, and even a quantitative relative qualification. Descartes in-
troduced space-referentials only in order to organize all such relative quanti-
tative characterizations, with respect to one common origin: a ‘legalization’
of reference!
Consider next the velocity of a particle. This mathematical concept
is by construction independent of the spatial coordinates, in the utilized
space referential, of the point where the velocity acts; it depends only on
the numerical values of differences between spatial coordinates. If the initial
referential is changed by translation, these differences do not change, though
their notational expression does. So, if the velocity vector were to change,
this would be logically inconsistent with the definition of the concept of ve-
locity. Again, to indicate the identified constraint in terms of ‘negative’ and
‘positive’ knowledge, seems somewhat miseading: the conservation of veloc-
ity with respect to translation of the space-referential (hence, in Newtonian
mechanics, also the conservation of momentum) is just a constraint of in-
ner logico-mathematical consistency of the system of the concepts involved,
namely space, space-referential, distance, velocity.
The general type of conceptual situation where considerations of sym-
metry are generative could reveal features that are much less immediately
intelligible in terms of constraints of, exclusively, logico-mathematical self-
consistency of the global mathematical representation. But in the trivial
cases just examined, instead of saying that what can be known has been
determined from what cannot be known, one can also only say that a math-
ematical descriptive concept cannot, without contradiction, be conceived to
change in consequence of changes of which, by construction, it is indepen-
dent. This is less mysterious, but also much less striking, of course.
Petitot. I don’t understand how in absolute space there cannot be
absolute positions. It is precisely when there is a relativity group acting tran-
sitively that all positions become equivalent. By definition, relativity negates
absoluteness. In fact, the problem is much deeper than a mere question of
logical coherence. What I have in mind is the determining role of Noether’s
MATHEMATICAL PHYSICS AND FORMALIZED EPISTEMOLOGY 83

theorem proving, in a variational context, the correlation between symme-


tries (negative geometrical limitations) and conserved quantities (positive
physical contents). We will return to this point later.
Mugur-Schächter. Please let me go back to ‘ontology’, to try to
come to a common conclusion. Is there some sort of ontology with which
you would finally agree?
Petitot. Well, I think that we have to radically reject in science any
ontology in the classical metaphysical sense. But there is another meaning
of the word “ontology”, philosophically tied with the concept of logical type,
and refering simply to the general categories of object, whole, part, relation,
set, function, event, process, etc., which we need to speak of something.
Here “ontology” means in fact “categorial frame”. It is the sense of the
term, e.g., in Husserl’s formal ontology, or in what is called in contemporary
AI “ontological design”.
Mugur-Schächter. We need? In what sense?
Petitot. In order to conceptualize the phenomena correctly.
Mugur-Schächter. Does ‘correctly’ mean so as to insure truth, or
intelligibility, or both?
Petitot. Both.
Mugur-Schächter. In short, if instead of ‘ontology’ I say ‘a cor-
rect model of intelligibility’ (to distinguish it from the purely mathematical
models’), I guess that we might finally agree. Only one more question in
this context: by intelligibility, would you mean availability of any model, or
only of a model that insures connection with the categories of objectivity in
a simple ‘explicative’ way?
Petitot. As I explained before, for me scientific intelligibility is es-
sentially linked with the possibility of an algorithmic simulation derived in a
principled way, or, to make use of another terminology, of a “computational
synthesis” correlated to a categorial analysis. Every time one succeeds to
produce a very economic computational synthesis rooted in general and fun-
damental principles, then intelligibility is achieved.
Bailly. In usual computation one allows entirely ad hoc models, un-
der strong technical restrictions, whereas, in the computation toward which
you point, the condition of a relation to first principles, to universal cat-
egories, is very important. You do not explicate this in your assertions,
which, in my opinion, weakens them.
Petitot. I strongly emphasized that the algorithms should be rooted
into objective categories and principles.
Bailly. Together with ‘simulation’ you should then also require this
explicitly and from the start, and declare the aim of intelligibility. But
instead of pursuing the discussion now, I think that we should let you first
84 J. PETITOT ET AL.

finish your exposition and continue afterwards on a more complete ground.


Petitot. OK. Let’s continue. I would like now to illustrate my
general purpose with a couple of examples.
The problematic of the forms of phenomenal givenness and of their
mathematization has been initiated by Kant in his Transcendental Aesthet-
ics. The interpretation of the categories of objectivity starting from the
“pure intuitions” is achieved at two distinct levels: first their “schematiza-
tion”, and then, at a much deeper level, their mathematical “construction”,
which radically transforms their traditional sense (therefrom the critical dis-
junction between scientific and common sense knowledge).
So, as we have already seen, in order to legalize phenomena, one
must use paradoxical objectifying tools which take into account how phe-
nomena hide their underlying being, and which, at the same time, bracket
any “psychological” processing. They must break with both ontology and
psychology.
Before considering the case of modern quantum microphysics, let me
briefly indicate how this fundamental problem has been solved by Kant
for classical mechanics in his First Metaphysical Principles of the Science
of Nature, so deeply analyzed by J. Vuillemin in his Kant’s Physics and
Metaphysics.
In the Phoronomics (Kinematics) Kant studies the categories of
Quantity and the associated transcendental principles called the “Axioms
of intuition” that rule the function of extensive magnitudes. Two problems
are treated there.
(a) The way in which measure becomes possible for the phenomenologi-
cal forms of space and time. The introduction of coordinates allows
an arithmetization of these continua, and then the introduction of a
metrics allows to measure distances. Thereby space as a continuous
medium for manifestation (“form” of intuition) becomes – in view of
physics – a geometry (“formal” intuition) whose Euclidean structure is
unseparable from mechanics (inertia principle and straight geodesics).
(b) The rectilinear uniform motions and Galilean relativity (that is the
invariance group of the theory). This refers on the one hand to space-
time symmetries: temporal translations, spatial translations and ro-
tations, etc. (Kant was the first philosopher to strongly stress that
spatial symmetries were constitutive of physical objectivity). On the
other hand this refers to the properly kinematic Galilean transforma-
tions (rectilinear uniform motions). Wherefrom the constitutive role
of the relativity principle. As J. Vuillemin notes,

“it is the relativity of motion which renders transcendentally


MATHEMATICAL PHYSICS AND FORMALIZED EPISTEMOLOGY 85

necessary the subjectivity of space (its transcendental ideal-


ity).”

It is noteworthy that for Kant the law of addition of velocities was by no


means obvious and even constituted a central problem. Indeed the veloci-
ties are intensive magnitudes, not extensive ones. So their additivity (their
vectoriality) was to be proven in agreement with their intensivity, which is
not a matter of course. Motion is not a mode of space, and the additivity is
not exclusively geometrical, but kinematical. We would say nowadays that
velocities belong to the tangent spaces of the ambiant space but that, due
to their vectorial structures, all these different spaces can be identified.
Kant’s treatment of Dynamics breaks with the Leibnizian view. For
Leibniz space was imaginary. The substantial interiority, the real substance-
force, was outside space, notwithstanding its spatial expression. Kant con-
serves this substantial interiority, but only as a (non-causal) metaphysical
foundation, which being of a noumenal nature cannot be a component of
objectivity and has to be determined exclusively via its exteriorization (its
spatio-temporalization through motion). So the problem is to work out a
purely spatio-temporal concept of Dynamics.
The categories of Quality and the associated principles called the
“Anticipations of perception” entail that matter can phenomenalize itself
only via intensive magnitudes like velocities and accelerations. But this
entails in its turn to root Dynamics into Phoronomics and, according to
J. Vuillemin, this is the true sense of the Kantian “Copernican” revolution
in philosophy. The systematic link between Dynamics and Kinematics is
expressed as a principle of covariance relatively to the Galilean invariance
group. After the transcendental interpretation of the relativity principle
in the Phoronomics, we attend there the transcendental interpretation of
another fundamental principle, namely that physical phenomena have to
be described by differential entities that vary in a covariant way. In short,
the Dynamics explains that, for transcendental reasons, Mechanics must be a
differential geometry (and not an Aristotelian logic of substances, properties,
and accidents).
So in the Dynamics the categories of quality become irreversibly di-
vorced from the traditional metaphysical concept of substance. Spacetime is
filled with matter. This filling-in which acts as a dynamical tension for oc-
cupation results from the conflict between attractive and repulsive internal
forces that generate the cohesion of bodies, their material phases and their
interactions. But these fundamental “primitive” forces – to be well distin-
guished from the “derivative external” mechanical ones – share a noumenal
being. The internal processes of matter which generate the dynamical qual-
ities remain outside any mathematical construction.
86 J. PETITOT ET AL.

It is only in the visionary and prophetic reflexions of the Opus Pos-


tumum that Kant tries to conceive of (through the concepts of energy and
aether) a fourth Critique, a Critique of the Physical Judgement which would
develop a new physics of the interiority of matter. Then matter will be re-
garded as that principle by what space becomes a sensible object that can
be given intuitively and empirically.
The categories of Relation (substance, causality, reciprocal action or
interaction) and the associated principles called the “Analogies of experi-
ence” correspond to the Mechanics part in the First Principles. In the
Mechanics, matter is reduced to mass, that is to a scalar number. This
allows to construct mathematically the concept of motion without having to
construct before that of matter itself. Mass becomes “the ultimate subject
in space”, motion becomes its “determining predicate”, and as far as this
predicate is purely spatio-temporal, matter is effectively treated on the basis
of its sole phenomenality. Matter stops to be a “second” physical matter
animated from beneath by an extra-spatial substantial materia prima. It
becomes a space-time-mass unity.
This objective conception of mechanics allows to construct mathe-
matically the physically fundamental “dynamical” categories of substance,
causality, and interaction.2 The construction begins with a complete reinter-
pretation of the traditional (Aristotelico-Scholastic) category of substance.
Already interpreted in the Critique of Pure Reason via the transcendental
temporal scheme of permanence, substance is now identified with the prin-
ciple of conservation of physical quantities, that is, with physical principles
of invariance. This reinterpretation is an epistemological breakthrough of
utmost importance. It breaks with all the previous logical and/or ontologi-
cal approaches that consider science as a predication on the states of affairs
of an independent reality. In Mechanics the category of substance becomes
the source of the conservation laws which, once translated into equations,
exhaust the content of the physical theories.
As to the causality principle, it is reinterpreted by the inertia principle
and Newton’s law.
Finally, the category of reciprocal action or interaction is reinter-
preted by the law of action/reaction equality and also, via the transcenden-
tal scheme of simultaneity, in terms of universal interaction. It is noteworthy
that simultaneity raised for Kant a very difficult problem. Indeed one has
to insure its objectivity. This requires a coordination of the local times into
a global cosmological time. Now, for Kant, the Universe was not an objec-
tive concept (see, e.g., the cosmological antinomies in the Transcendental
2
For Kant, the “dynamical” categories of Relation are the true physical ones, as opposed
to the “mathematical” categories of Quantity and Quality.
MATHEMATICAL PHYSICS AND FORMALIZED EPISTEMOLOGY 87

Dialectics). It was only a regulative Idea.


The categories of Modality (possible, real, necessary) and the associ-
ated principles called the “Postulates of empirical thinking in general”, are
treated in the Phenomenology part of the First Principles according to their
definition in the first Critique:
1. what agrees with the formal conditions of experience (pure intuitions
and categories) is possible;
2. what agrees with the material conditions of experience (sensation) is
real;
3. that whose agreement with the real is determined according to the
general conditions of experience is necessary.
It is essential here to stress that the Kantian concept of reality is a modal
category which has an objective meaning only relatively to constitutive pro-
cedures. Due to relativity, motion cannot be a real predicate, but only a
possible one. It cannot be regarded as a real change of the real inner state
of the system and of some of its intrinsic mechanical properties. By reduc-
ing matter to mass, by rejecting “primitive internal” forces and considering
exclusively “external derivative forces”, Mechanics cannot draw from mea-
surements on motion any determination of the inner state of the system.
Therefore, one can at the same time assert and negate motion without any
logical contradiction. In other terms, relativity of motion renders unaccept-
able the surreptitiously ontological interpretation of statements like “the
body S has that position or that velocity”, were “to have” would mean “to
possess a property”. Such statements do not support counterfactuality since
their truth value presupposes that an inertial referential has been chosen
(i.e. that the conditions of measurement have been fixed). For Kant, there
exists therefore an irreducible conflict between physical objectivity and com-
mon sense natural (predicative) logic. Of course it remains correct to speak
“as if” localization and motion were “properties” of bodies. But this naive
“empirical realism” of space can in no way be scientifically interpreted as a
“transcendental realism”. The celebrated thesis on the “transcendental ide-
ality” of space only expresses in philosophical terms this modal consequence
of the relativity principle.
However Dynamics does provide criteria of reality for motion, since
Newtonian forces are invariant relative to the Galilean transformations and
are therefore real.
One sees to what a point the elimination of the transcendental per-
spective by logical empirism has been ruinous for the philosophy of physics.
Kant made an outstanding effort in order to clarify the fundamental epis-
temological problems of physical objectivity: the opposition between objec-
tivity and ontology, the prescriptive character of the categorial legalization,
88 J. PETITOT ET AL.

the constitutive role of symmetries, conditions of covariance, and conserva-


tion laws, the modal character of reality, the inadequation of logic due to
its “dogmatic” nature, etc. But instead of continuing his effort in parallel
with scientific progress, philosophers rejected it for dogmatic reasons and
came back to a scholastic logicism correlated with an ontology clearly in-
compatible with physical objectivity. Fortunately, the physicists themselves
retrieved the genuine sense of the transcendental approach to scientific ex-
perience and of the critical elements of knowledge: they taught philosophers
a good lesson!
I would like now, before the final discussion, to conclude with some
remarks concerning QM.
We have seen that, in order to legalize phenomena one must use of
a paradoxical objectivizing instance, which breaks at the same time with
ontology and with psychology. We also have seen how this fundamental
problem has been solved in classical mechanics via the concepts of space
and time. I hold that in QM it was solved via the concept of probability
amplitude. Indeed, this constitutive concept allows to take into account the
inseparability between the micro objects and the measuring devices, without
having to bring into play any particular theory of any specific apparatus.
This fact determined the choice of a new type of mathematics for interpreting
the objective categories, and in particular the “dynamical” (physical) ones.
So, I shall argue that in QM probability amplitudes play a transcen-
dental role analogous to that played by space, time, and differential calculus
in classical mechanics. I recall the four characteristic properties of probabil-
ity amplitudes (PAs) (|ii = initial state, |f >= final state):

1. If there exist k indiscernible transition paths, then the corresponding


PAs behave additively:
X
hf |ii = hf |iik .
k

2. If there exist several discernible final states |f ik , then the PAs behave
additively:
X
|hf |ii|2 = |hf |iik |2 .
k

3. If the transition |ii → |f i is achieved through an intermediary state


g, then there is a factorization of the PAs:

hf |ii = hf |gihg|ii.
MATHEMATICAL PHYSICS AND FORMALIZED EPISTEMOLOGY 89

4. If there are several independent systems, then there is again a factor-


ization of the PAs:

hf1 f2 |i1 i2 i = hf1 |g1 ihf2 |g2 i.

These axioms are to QM what those of Euclidean geometry are to classical


mechanics. If the paths are interpreted as classical trajectories, they lead
immediately to Feynman’s path integrals. On this basis, QM can derive its
mathematical models in much the same way as classical mechanics derives
its mathematical models from space-time geometry and differential calculus.
In QM probability amplitudes express potentialities actualized by measure-
ments. The relational nature of the concept of quantum state (often stressed
by Bohr), as well as its interpretation in terms of a relativity principle, have
been strikingly well formulated by Vladimir Fock:

“The probabilities expressed by the wave function are the probabil-


ities of some result of the interaction of the micro-object and the
instrument (of some reading on the instrument). The wave function
itself can be interpreted as the reflection of the potential possibili-
ties of such an interaction of the micro-object (prepared in a definite
way) with various types of instruments. A quantum mechanical de-
scription of an object by means of a wave function corresponds to
the relativity requirement with respect to the means of observation.
This extends the concept of relativity with respect to the reference
system, familiar in classical physics.”3

So in QM, transcendental aesthetics has undergone a mutation. It is no


longer perceptively based, but purely instrumental. Its critical role is how-
ever the same, and many over-elaborate philosophical discussions concerning
QM stem precisely from the difficulty to understand the strictly objective
character – neither ontological, nor subjective, nor classically statistical – of
the indeterminism tied with the concept of probability amplitude.
If we compare now classical and quantum mechanics, we can con-
clude that the function of a “transcendental aesthetics” in a procedure of
constitution is characterized by the four following requirements:
(i) to determine forms of manifestation (general conditions of observabil-
ity and universal formats for informations) that allow us to take into
account the relational status of phenomena while bracketing at the
same time their “subjective” internal content;
(ii) to define relativities that violate the principles of any substantial on-
tology;
3
Quoted by Max Jammer.
90 J. PETITOT ET AL.

(iii) to provide a mathematical basis for the mathematical construction


of the “dynamical” (physical) categories of substance, causality, and
interaction;
(iv) to lead to an interpretation of the modal categories of possibility (po-
tentiality, virtuality), reality (actuality), and necessity which are no
longer absolute (metaphysical) but only relative to the constitution
procedure (physical).

This last point is crucial from an epistemological point of view. Indeed,


the main philosophical problems raised by physical realism stem from a
misunderstanding of the modal character of reality as a category. Modality
implies that some objective “properties” do not support counterfactuality
and, without any logical contradiction, can be at the same time asserted
and negated (they cannot therefore be possessed by substantial individuated
objects, in the ontological sense of the verbs “to possess” or “to have”).
Bitbol. I remember that for you ‘weak objectivity’ in the sense of
d’Espagnat, is a form of recognition of the distinction which you rightly make
between objectivity and ontology of an independent reality. But d’Espagnat
answers to this that in classical physics, even though philosophically the
objectivity defined there indeed cannot be identified with an independent
ontology, it is nevertheless possible to speak and think as if one were in
presence of a description of an independent reality...
Petitot. Yes, but classical physics can’t concern a fully independent
reality because of Galilean relativity.
Bitbol. ... whereas according to d’Espagnat quantum mechanics is
much more radically in a different position. Starting from quantum mechan-
ics it seems much more difficult to imagine an ‘independent ontology’, if not
altogether impossible. So you are quite right to point out the fine clues
which already in classical mechanics led toward this idea that no indepen-
dent ontology is conceivable.
Petitot. Of course, Bernard d’Espagnat is, as always, perfectly right.
The ontological “crisis” is very much stronger in QM. But, in my opinion,
this is so in part because we are modern minds totally immersed at the
onset into the evidences of classical physics. We do not even remember
how Aristotelians used to think. We are all naively Galilean. We have
forgotten that the discovery of relativity was a true trauma, even long after
Galileo. What was at stake was the understanding that “having” this or
that position and this or that velocity is not a substantial predicate, and
cannot be assigned as a property to the material bodies. Van Fraassen says
that in QM “states can be identified in terms of observables, but cannot be
identified with them”. It is already the same thing in classical mechanics:
a moving object can be identified in terms of position and velocity, but its
MATHEMATICAL PHYSICS AND FORMALIZED EPISTEMOLOGY 91

internal state cannot be identified with these dynamical descriptors.


Bailly. Certainly. But if you make use of a referential, you can:
everything, then, is transferred in the referential relation.
Petitot. This is the relativity principle. As Fock emphasized, in this
respect QM generalizes classical relativity.
Bailly. Yes. Concerning relativity I have a tendency to speak in
terms of a sort of ‘order’, nearly an ‘approximation process’: Aristotelian
theory was at the order zero, Galilean theory was at the first order, general
relativity is at the third order, and quantum mechanics is a sum of all this
up to infinity.
Mugur-Schächter. To go to the bottom of this line of thought,
one should fully realize that no qualification whatever is ‘predicable’ in an
absolute way. Any qualification presupposes some sort of point of view,
so some sort of referential. A predicate, quite essentially, is a referential
relation. No qualification can be conceived as ‘a property of the qualified
entity’. This expression is either a shorthand expression for ‘a property of
. . . with respect to . . . ’, or, if not, it is a self-contradictory concept.
This is obscured by two facts. First, that very few classes of refer-
entials have been reconstructed mathematically, like those from nowadays
physics. (For instance, the referentials for the qualifications ‘blue’, ‘big’, ‘in-
telligent’, etc., are not even only formally constructed – in a ‘legalized’ way
– so far they remained vague and implicit.) And, second, that sometimes
the locution ‘to have something ’, where ‘thing’ stays for a substantive, is
globally assimilated with a predicate, on the basis of linguistic characters
(for instance, grammatically, ‘position’ is a substantive, so one tends to as-
similate ‘to have positions’, with ‘to have points’, like absolute space, and
to treat this like a predicate). A mixture between grammar and logic.
Petitot. Quite so. I agree. “To have a property” does not support
conterfactuality and therefore there exists an irreducible conflict between
physical objectivity and natural logic. Physics is not predicative in that
sense.
Mugur-Schächter. Natural logic is not special in this respect. As
far as I know, a logic cleaned of any trace of absolute predication does not
yet exist? A fortiori there does not yet exist a fully relativized logic, taking
into account – besides the relativity to the referential, involved in predica-
tion – also the unremovable relativity of any description (representation), to
the process of generation of the described entity. This last sort of relativity
is particularly obvious in quantum mechanics; but, more or less explicitly
individualized with respect to predication, it is present in any process of
representation. Nevertheless, concerning quantum mechanics as well as in
general, it is still very weakly perceived. A certain type of analysis of quan-
92 J. PETITOT ET AL.

tum mechanics, if it is sufficiently deep, leads rather directively toward a


full relativization of logic.
Bitbol. I would like to go over to another question. You say that
objectivity in Kant’s sense is a legalization of phenomena. I am not certain
that this applies well to quantum mechanics. The laws of quantum me-
chanics, the equations of evolution, etc., apply to abstract objects, to state
vectors in a Hilbert space, not to an entity immersed in physical spacetime
where the phenomena are realized. Do you then think that this expresses
an even more profound translation than that concerning ontology, to pass
from classical physics to quantum mechanics?
Petitot. This, indeed, is a key point. In what I presented here,
the term “phenomenon” meant “observable” and not merely “sensible phe-
nomenon”. I have already stressed that physics deals only with phenom-
ena and that the axioms of a physical theory must include their relational
essence. Now, how do the axioms of QM realize this condition? By sub-
stituting the concept of “observable” for that of sensible phenomenon. An
observable is not a mere brute phenomenon, nor a phenomenon prepared in
a definite way. It is already a pre-objectivized structure. Here appears a
quite essential difference between classical and quantum mechanics. Classi-
cal mechanics is a theory of motion as pure exteriority. As we have seen,
mechanical forces are “derivative” secondary forces produced by a noume-
nal “interiority” of matter on which classical physics can say nothing. If
the observed phenomena are pure motions, then to convert them into ob-
servables means essentially to geometrize them by introducing spacetime,
trajectories, and equations whose the latter are solutions. As we have seen,
matter is reduced to mass and intensive magnitudes like velocities and accel-
erations. At each point of spacetime there are dynamical vectors, and what
has to be reconstructed by the physical laws are the fields of such vectors.
So the corresponding physical theories will necessarily be ruled by systems
of differential equations.
From the transcendental point of view, QM represents an extraordi-
nary comeback of “interiority”. It introduces the presupposition that the
interiority of matter can be described by internal quantum numbers. The
Hilbert space of the states of a quantum system is in general not completely
determined. Its relation to spacetime is not necessarily explicit. The only
thing we know is that it must exist and that it is a good frame for mathema-
tizing the results of measurements. But in fact the fundamental quantum
structure is the algebra of observables. The Hilbert space can be retrieved
from it. Moreover, the quantum internal degrees of freedom can depend on
space-time and can be geometrized as in gauge theories by fibre bundles.
Then the Hilbert space of quantum states becomes much more explicit geo-
MATHEMATICAL PHYSICS AND FORMALIZED EPISTEMOLOGY 93

metrically: it is the space of L2 sections (or more precisely of L2 densities).


Bailly. There even exist a sort of geometrical aspect of interiority,
for example in the representation of the effects concerning the detection of
vector potentials. This interiority of which you speak manifests itself not
only on the level you just mentioned, but even in the relation between the
structure of physical spacetime and the property of phase, for instance.
Petitot. And if you consider (super)-string theory, the physical
spacetime position becomes a sort of “quality” of the string. The true quan-
tum object is an abstract non-embedded string, and spacetime coordinates
correspond to functions (observables) defined on it.
Bailly. That is why the theory is compatible with gravitation.
Bitbol. Some authors assert that in quantum mechanics the unique
space of objectivity is the Hilbert space. It simply is exlusively there that
one can make operate the categories of causality, invariance, etc.
Petitot. Absolutely. And if one goes further and drops the Hilbert
space, conserving only the algebra of observables, the whole categorial struc-
ture can be anew interpreted (mathematically “constructed”) in this opera-
tor framework. This raises a beautiful question of formal epistemology: what
are the formalisms able to yield a mathematical semantics to the categorial
syntax? There exist very few formal universes where a categorial syntax can
be coherently interpreted. It is not at all evident to take for instance the
category of causality and to find a mathematical universe where it becomes
possible to say: “causality means this or that”. Physicists succeeded in real-
izing this outstanding performance for classical mechanics using Euclidean
geometry and differential calculus, and for QM using Hilbert spaces and von
Neumann algebras of operators, but these achievements are masterpieces of
theoretical invention!
Bitbol. Finally, however, this raises questions. What is finally the
material object of quantum mechanics? If the object is determined by the
net of legal reconstruction of the phenomena, where it is inserted by the
theory, then the ‘place’ of this reconstruction defines the nature of the con-
sidered object. But with respect to Kantian categories that act in the phys-
ical spacetime the object is still material, even though it is not perceptible.
It certainly is no more possible to imagine this material object in classical
terms, as a resistant extended but confined entity, etc. So how could it be
conceived?
Petitot. That is difficult to say. The quantum void?
Bitbol. Yes, the quantum void: this would lead perhaps to regard
the material object of quantum mechanics in terms of propensities.
Mugur-Schächter. The quantum void cannot play the role of an
‘object’ in Kant’s sense, in any case not of a primary material ‘object’: it is
94 J. PETITOT ET AL.

not a phenomenon, and a fortiori not a legalized one. (In Opus Postumum
Kant considers the ‘ether’ to be ‘a necessary postulate’.)
For the same reason, the ‘states of microsystems’ either cannot be
the Kantian material ‘object’ of quantum mechanics, even though – as the
supposed result of a definite operation of state-generation – each such state is
a ‘legalized’ entity, which furthermore is precisely what quantum mechanics
basically ‘describes’ ! (The ‘objectual’ characters of microsystems, mass,
spin, etc., can be regarded as invariants with respect to changes of ‘state’).
As to the representations of the physical states of microsystems by
a Hilbert state vector, these being formal entities, they cannot do either.
Furthermore, though in relation with quantum mechanics it is often spoken
of the ‘individual phenomena’ – that is, one single observable mark on an
apparatus, produced by only one realization of a measurement on some
previously generated state – these, though they are phenomena, are not
legalized quantum mechanical phenomena in Kant’s sense.
So Kant’s definition of the ‘object’ appears to be pretty restrictive.
The only more resistant candidate, I think, are [the probabilities of realiza-
tion of the various observable marks produced on an apparatus in conse-
quence of a succession of two operations – an operation of state-generation
followed by a measurement operation – repeated a number of times suffi-
ciently big for permitting to a probability law, if it exists, to manifest itself].
All this together. Strictly speaking, these probabilities also are far
from being ‘direct’ phenomena: in contradistinction to the individual marks
that are mathematically represented by the quantum mechanical eigenvalues
of operators, the mentioned probabilities cannot be observed, they have to be
calculated on the basis of countings of the registered marks and accordingly
to theoretical specifications. However they are the descriptional entities from
quantum mechanics that are the most immediately connected with directly
observable phenomena, and they are ‘legalized’.
But the most striking is this: In the case of quantum mechanics, there
is no initial phase of spontaneous ‘donation’ of forms of phenomenal manifes-
tation. In so far as one can speak of phenomenal manifestations, the ‘forms’
of these are in fact the probabilistic meta –forms just mentioned, and these
emerge only already legalized by the quantum mechanical normed procedure
for the generation of the ‘objects’; they are these ‘objects’, directly: so these
forms either are actively constructed for the definite purpose of qualifying
the hypothetical ‘states of microsystems’, or they simply do not exist.
Under these conditions, what, exactly, becomes of ‘transcendental de-
duction’ ? It starts from what? From just the aim to generate knowledge
concerning hypothetical, non-phenomenal entities. In this situation it might
be fruitful to reconstruct from A to Z a modern version of the Kantian defini-
MATHEMATICAL PHYSICS AND FORMALIZED EPISTEMOLOGY 95

tions, in order to clearly include the quantum mechanical approach (as well
as the other modern theories). This might be a really interesting purpose on
the way toward a formalized epistemology. It might appear appropriate, in
such a modernization, to introduce also certain modifications of the Kantian
terminology. For instance, personally, instead of ‘dynamical categories’, I
would prefer to say ‘categories of intelligibility’.
Petitot. There are many different questions in your remarks. First
of all, I try to elaborate a transcendental epistemology of modern physics
and of course not to project Kant on modern physics. There is something
very strange with Kant. When Hilbert applied to modern mathematics the
axiomatic perspective, nobody said he wanted to reduce mathematics to
Euclid. When Van Fraassen claims that he works out an empirist episte-
mology of QM, nobody says he wants to reduce QM to Hume. It might
be the same thing with the transcendental approach. It comes from Kant
but it is not a reduction to Kant. So I completely agree with the idea that
it would be interesting to reconstruct a modern version of Kant including
QM. My agreement is even so strong that this is in fact my research program
since many years. I have already writen a lot on these subjects, given many
invited talks in a lot of Conferences and Symposia, and devoted to it many
seminars.
One of the main problems we meet in generalizing the transcenden-
tal approach is to avoid any surreptitious regression to a naive, precritical,
conception of phenomena. Concerning QM, I have strongly emphasized that
quantum “transcendental aesthetics” is provided by probability amplitudes.
So I agree with you: in QM the “forms of givenness” are in some sense
probabilistic “meta-forms”. But it is not so evident that they are really
meta-forms. For it is wrong to think that our naive perception is immediate
and direct. In fact, the visual system is a physically hypercomplex apparatus
of measurement, information processing and cognitive representation. For
instance the retina is a fantastic system of correlated quantum photorecep-
tors. Moreover the interpretations of images processed by the visual cortex,
are bayesian processes extracting statistical regularities from sensorial data.
So a classical phenomenon is not so “classical” than it can seem. Moreover,
in classical mechanics also phenomena are not direct: you need instruments
(Galileo’s telescope), preparation (Galileo’s inclined plane), changes of ref-
erentials (Copernican heliocentric system), etc., etc., and the phenomena
are also actively constructed and theoretically laden.
In what concerns now the fact that quantum states cannot be Kantian
objects, it is trivial. Quantum states correspond to dynamical states in
classical mechanics, that is, positions and velocities, or, as we say nowadays,
points in phase space. They correspond to “mathematical” categories, and,
96 J. PETITOT ET AL.

as I said, the truly objective categories are the “dynamical” ones. The object
must be viewed as the correlate of the “construction” of these categories.
For that, we need sophisticated mathematical tools, conservation laws and
Noether’s theorem (substance), forces (causality) as they are geometrized
in general relativity or QFT, interactions as they are geometrized in gauge
theories, etc.
In what concerns finally your lexical suggestion for “categories of
intelligibility”, I think that the crucial problem is not intelligibility but ob-
jectivity and reality.
Bailly. Do you think, Jean, that we should distinguish between
building an analysis of physical theories, and the construction of a formalized
epistemology?
Petitot. This is an important question. I would say that a for-
mal epistemology must concern primarily the mathematization procedures.
However, this does not exhaust the problem. For two reasons. First, be-
cause, on the side of mathematics, we need an epistemology of the mathe-
matical structures themselves and of their relations with physical objects.
And second, because, as we have seen repetedly, to be explanatory any al-
gorithm has to be connected to the categories and principles of objectivity.
It would be perhaps useful to root the approach into a “theory of the ob-
ject in general”, a “formal ontology” in Husserl’s sense. But there the term
“ontology” is extremely weak. It is in fact semantic and has to be related
with our cognitive processes.
But it is true that I consider that even these representations must
at the end become mathematical. Indeed, what are the most achieved at-
tempts at a formal ontology? First set theory, and second category theory.
Therefrom stem modern logic, model theory (relations between syntax and
semantics), etc. All this is mathematical.
In short, a first aspect of a formal epistemology should deepen the
concept of a general object as a mathematical construct in relation to
physics. A second one should deepen it on the side of categories and princi-
ples. And a third one should bring all this in relation with cognitive sciences.
Mugur-Schächter. The way of beginning, and the order of progres-
sion – which can violate certain ‘natural’ orders – is very important. I think
that a ‘general theory of objects’ can only be a general representation of
normed descriptions, since nothing else than descriptions exists that can be
both ‘known’ and communicated. Legalized physical phenomena are normed
descriptions of physical data, legalized concepts are normed descriptions of
abstract data, legalized psychological phenomena.
Bitbol. Indeed, a formal epistemology could first deal with the gene-
sis of legalized phenomena, with the process of constitution of a ‘procedural
MATHEMATICAL PHYSICS AND FORMALIZED EPISTEMOLOGY 97

rationality’. It could first concentrate upon the genesis of the agreement


upon phenomena, which later can be reconsidered inside a further more
general representation. Does there not exist something that is subjacent
to the perfectly organized rationality which works in mathematics and in
axiomatized science? Groping experimentation, comparisons....
Petitot. Certainly.
Mugur-Schächter. We should methodologize all this.
Bitbol. That would be the underground of the formal epistemology.
Mugur-Schächter. Its basis.
Petitot. To conclude, I would like to add a few remarks on the
specific contents of the main categorial moments in modern physics and on
how they could drive the research programm of a formal epistemology.
Kant studied the categorial moments of classical mechanics. We have
to continue his transcendental research programm. We can first examine the
categorial moments of the classical theory of fields (the theory of continuous
media from the beginning of the last century up to now). We can then do
the same for general relativity and QM. I think that I have already shown
that it is possible.
Secondly, we can analyze the transcendental trend that drives the
progressive geometrization of physical theories. In classical mechanics force
is the physical expression of the categories of causality and reality. It is cat-
egorialy “dynamical” and not “mathematical”. Forces are invariant with re-
spect to the Galilean relativity group and are therefore endowed with a phys-
ical reality. In general relativity, the physical content assigned previously
to forces, is transferred into the (pseudo-Riemannian, locally Lorentzian)
metric of space-time. The symmetry group of the theory becomes incredi-
bly larger (it is now the infinite dimensional Lie group of diffeomorphisms)
and, as a consequence, all the forces become “inertial”. This means that
Mechanics is absorbed in a generalized Kinematics. In terms of the cat-
egorial structure, it is as if causality and reality were converted into pure
“phoronomical” a prioris. This is possible because, conversely, the previ-
ously purely “mathematical” moment of the metric of space-time (Axioms of
intuition) is converted into a metric moment endowed with physical content.
The physical categories of causality and reality are lifted up into the geo-
metrical ones, whereby correlatively geometry becomes laden with a physical
content. By this sort of double movement the structure of the categorial sys-
tem is completely changed. But of course it is still constitutive. In fact, as
was deeply stressed by Ernst Cassirer, general relativity is one of the most
achieved examples of transcendental physics. In a similar way, Noether’s
theorem, via the principle of least action, links the symmetries of a theory
to its physical contents. This can be applied as soon as a variational formu-
98 J. PETITOT ET AL.

lation is available. This fact expresses the Galoisian quintessence of physics:


what can be known and measured is determined from what cannot. I think
that the philosophical deepness of this theorem, is far from having been fully
recognized and understood.
Bailly. What is needed in Noether’s theorem is just a metric, since
a geodesic is researched. The approach is equivalent to that one which you
mentioned before.
Petitot. Not really. You need only:

(i) the canonical symplectic structure of the cotangent bundle of the con-
figuration space,
(ii) a symplectic group action on it, and
(iii) the invariance of the Hamiltonian.

And again there is a deep transformation of the categorial structure. The


category of substance, already converted in conservation laws, becomes now
a direct consequence of symmetries. Once again, the “mathematical” (ge-
ometrical) categories become physically laden, while at the same time the
“dynamical” ones are geometrized.
Gauge theories deepen still this trend. Hermann Weyl, who was
not only a giant of mathematics and physics (inventor of the concept of
gauge symmetry) but also a specialist of transcendental philosophy and phe-
nomenology, said explicitly that his aim when creating the gauge concept
was to transform relativity principles into dynamical ones, that is to iden-
tify enlarged groups of symmetries with principles endowed with physical
content.
This is a sort of “teleological” trend toward a unification of the cate-
gorial moments. The rigid categorial Kant’s hierarchy is no longer operative,
according to which one first specifies the geometrical framework, then pre-
scribe a prioris upon what happens inside the framework, then builds laws.
In modern physics the categorial moments are at the same time “mathemat-
ical” (geometrical) and “dynamical” (physical). This is the transcendental
sense of geometrization. As a consequence, physical contents can be more
and more reduced to enlarged symmetry groups and generalized relativity
principles.
Mugur-Schächter. I suppose that in every new investigation one
has to achieve first the previous phases. One cannot find directly the most
economic algorithms. These can be identified only in the end. But they
certainly deserve a thorough investigation. As also does the whole previous
genesis. Without an explicit and methodologized knowledge of the genesis
one cannot methodically achieve intelligibility, I think.
So, your program, Jean . . .
MATHEMATICAL PHYSICS AND FORMALIZED EPISTEMOLOGY 99

Petitot. . . . is to revisit the difficult problems of categorial structure


and to work out a series of detailed case studies. To begin with “exercises
of applied epistemology”, then to elucidate the way in which, in each case,
the categorial structure is mathematically schematized and constructed, and
finally to show how the computational synthesis of phenomena can be algo-
rithmically deduced from such constructions.
Bailly. Do you not think that the movement of geometrization that
you so well described, might lead to confusion concerning our relation to
the world, since the human need for explanation is tied to the ‘natural’ dis-
tinction between categories, such as Kant described them. Would you think
that there emerges now a new sort of explanation, in terms of unification
of the categories in structures of cognition, that will be able to reconstruct
otherwise all that has been done before in philosophy in terms of separate
categories?
Petitot. I don’t know how to answer your question. My second
domain of research deals with human cognition conceived of as a natural
phenomenon. On the other hand, I have explained how and why my episte-
mology of physics is transcendental. Of course, the time must come, some
day, for working out a relation between transcendental epistemology and
natural cognition. But for the moment I make strong reserves concerning
a naturalized epistemology of mathematical physics. I fear a vicious circle.
Cognition is a natural neuro-biophysical phenomenon to which it is pos-
sible to apply sophisticated mathematical models. On the other hand, my
epistemological views are strongly focused on the problem of mathematizing
physical phenomena. If I tried to apply directly cognitive models to them,
I would need, to be consistent, a cognitive theory of mathematics. But how
could such a theory be obtained? Only by the help of mathematical models
of the brain. This is a vicious circle.
Mugur-Schächter. I think that it is not possible to have in advance
a general scheme for a not-yet-achieved development. One has just to begin,
hoping that the development will progress along a spiral, and will also permit
reflexive returns along the spiral, but without ever leading to circular face
à face collisions on a plane level.
Bitbol. There always are blind spots in knowledge, which change
with knowledge.
Petitot. Yes. This reserve is for me an only temporary one.
Mioara Mugur-Schächter. In any approach one has to start from
posited concepts and assertions. At the basis of a representation of episte-
mology, these a prioris might always remain different from those needed for
a representation of cognition regarded as a biophysical phenomenon. Be-
tween these two sorts of representations there probably acts like a sort of
100 J. PETITOT ET AL.

two-mirrors relation.
Petitot. Perhaps. This reminds me of a remark by Husserl: he said
the categories of our living world and those of the physical objectivity share
the same names, but are nevertheless completely different because the first
ones are not concerned with any mathematical “substruction”. This remark
is very deep indeed.
Bitbol. I propose a metaphor: The mirrors are not exactly parallel.
Mugur-Schächter. Well, we all thank you, Jean, very much indeed.
Petitot. I thank you all. That was really a very kind and exciting
exchange, and I want to thank particularly Mioara for having organized it.

REFERENCES

1. R. Abraham and J. Marsden, 1978, Foundations of Mechanics (Ben-


jamin Cummings, New-York).
2. H. E. Allison, 1983, Kant’s Transcendental Idealism. An Interpretation
and Defense (Yale University Press, New-Haven).
3. V. Arnold, 1976, Méthodes mathématiques de la mécanique classique
(Mir, Moscow).
4. D. Bailin and A. Love, 1994, Supersymmetric Gauge Field Theory and
String Theory (Institute of Physics, London).
5. D. Bennequin, 1994. “Questions de physique galoisienne,” in Passion
des Formes, à René Thom, M. Porte, ed. (E.N.S. Editions, Fontenay-
Saint Cloud), pp. 311-410.
6. M. Bitbol, 1996, Mécanique quantique: Une introduction philosophique
(Paris, Flammarion).
7. N. Bohr, 1935, Phys. Rev. 48, 696.
8. G. Brittan, 1978, Kant’s Theory of Science (University Press, Prince-
ton).
9. E. Cassirer, 1918, Kants Leben und Lehre; translation by J. Haden,
1981, Kant’s Life and Thought (Yale University Press, New Haven).
10. G. Cohen-Tannoudji and M. Spiro, 1986, La Matière-Espace-Temps
(Fayard, Paris).
11. H. Duncan, 1984, “Inertia, the communication of motion, and Kant’s
third law of mechanics,” Phil. Sci. 51, 93-119.
12. B. d’Espagnat, 1985, Une incertaine réalité (Gauthier-Villars, Paris).
13. B. d’Espagnat, 1994, Le réel voilé (Fayard, Paris).
14. H. J. Folse, 1978, “Kantian aspects of complementarity,” Kant-Studien
69, 58-66.
15. R. J. Gomez, 1986, “Beltrami’s Kantian view of non-Euclidean geom-
etry,” Kant-Studien 77 (1), 102-107.
MATHEMATICAL PHYSICS AND FORMALIZED EPISTEMOLOGY 101

16. J. Honner, 1982, “The transcendental philosophy of Niels Bohr,” Stud.


Hist. Phil. Sci. 13, 1, 1-29.
17. C. Itzykson and J. B. Zuber, 1985, Quantum Field Theory (McGraw-
Hill, New York).
18. M. Jammer, 1974, The Philosophy of Quantum Mechanics (Wiley, New
York).
19. M. Kaku, 1988, Introduction to Superstrings (Springer, New York).
20. I. Kant, 1781-1787, Kritik der reinen Vernunft (Kants gesammelte
Schriften, Band III, Preussische Akademie der Wissenschaften (Georg
Reimer, Berlin, 1911).
21. I. Kant, 1786, Metaphysische Anfangsgründe der Naturwissenschaft
(Kants gesammelte Schriften, Band IV, Preussische Akademie der
Wissenschaften (Georg Reimer, Berlin, 1911); translation by J.
Gibelin, 1971, Premiers Pincipes métaphysiques de la Science de la
Nature (Vrin, Paris).
22. I. Kant, 1796-1803, Opus Postumum, translation by F. Marty (Presses
Universitaires de France, Paris, 1986).
23. Y. Manin, 1988, Gauge Field Theory and Complex Geometry
(Springer, New York).
24. J. Marsden, 1974, Applications of Global Analysis in Mathematical
Physics (Publish or Perish, Berkeley).
25. C. W. Misner, K. S. Thorne, and J. A. Wheeler, 1973, Gravitation
(Freeman, San Francisco).
26. R. Omnès, 1994, Philosophie de la science contemporaine (Gallimard,
Paris).
27. J. Petitot, 1991, La Philosophie transcendantale et le problème de
l’Objectivité (Entretiens du Centre Sèvres), F. Marty, ed. (Osiris,
Paris).
28. J. Petitot, 1992a, Physique du Sens (Editions du CNRS, Paris).
29. J. Petitot, 1992b, “Actuality of transcendental aesthetics for modern
physics,” 1830-1930, in A Century of Geometry, L. Boi, D. Flament,
and J.-M. Salanskis, eds. (Springer, New-York).
30. J. Petitot, 1994b, “Esthétique transcendantale et physique mathématique,”
Neukantianismus: Perspektiven und Probleme, E. W. Orth and H.
Holzhey, eds. (Königshausen & Neumann, Würzburg), pp. 187-213.
31. A. Philonenko, 1972, L’oeuvre de Kant (Vrin, Paris).
32. C. Quigg, 1983, Gauge Theories of the Strong, Weak, and Electromag-
netic Interactions (Benjamin-Cummings, Reading, MA).
33. J. M. Souriau, 1975, Géométrie symplectique et physique mathématique
(Collection Internationale du CNRS 237, Paris).
34. B. C. Van Fraassen, 1991, Quantum Mechanics: An Empirist View
(Clarendon, Oxford).
102 J. PETITOT ET AL.

35. J. Vuillemin, 1955, Physique et Métaphysique kantiennes (Presses Uni-


versitaires de France, Paris).
36. A. Weinstein, 1977, Lectures on Symplectic Manifolds (Conference Se-
ries, American Mathematical Society 29, Providence, RI).
37. C. F. von Weizaäcker, 1979, Die Einheit der Natur (Hauser, Munich).
38. H. Weyl, 1922, Space-Time-Matter (Dover, New-York).
39. J. E. Wiredu, 1970, “Kant’s synthetic a priori in geometry and the rise
of non-euclidean geometries,” Kant-Studien 61 (1) 5-27.
6
ON THE POSSIBILITY OF A FORMALIZED
EPISTEMOLOGY∗

Robert Vallée

Université Paris-Nord
Paris, France

We present an attempt to a mathematical epistemology valid at the macro-


scopic level. Do similar perceptions correspond to similar objects? How can
we recompose two perception processes? The use of mathematical observa-
tion operators provides an answer. Identical perceptions may correspond to
different objects perceived, due to the lack of an inverse of an observation
operator. Consequently, there is a process of inverse transfer of structures
inherent to the observing system on the observed world. Moreover, multi-
plication of observation operators gives a representation of the composition
of perception processes. So we have at hand an algebra opening the way
to a mathematical and thus formal epistemology. Also, the intervention of
decision, composed with perception, allows the introduction of considera-
tions analogous to those valid in the purely epistemological case, giving rise
to a formal epistemo-praxiology. The possibility of a formal epistemology,
even completed by praxiology, at the macroscopic level provides arguments
in favor of a general formal epistemology acceptable at all levels.

Key words: algebra, epistemo-praxiology, formal epistemology, inverse


transfer, mathematical epistemology, observation operator.

On the paths of knowledge, for which there is no royal way, or maybe on


the banks of the unknown, the human mind has met two temptations: di-
rect, inexpressible apprehension and systematic, formal process. Between
these opposite poles, these two attractors, of enlightenment and some “ars
magna”, mind is still hesitating. But if the first type of knowledge is strictly
individual and subjective to an extreme, the second one, on which focuses
here our interest, should be valid for all. Is the attractor corresponding to
the second type so strange that it leads just to a mirage or, on the con-

See “Important Note” on p. xviii.

103
104 R. VALLÉE

trary, does it indicate a goal really accessible or which we can approach


asymptotically?
As a preliminary, it may be interesting to consider the acquisition of
knowledge, or much more simply the process of perception, at the macro-
scopic level. If this non-trivial process seems, at least partly, formalizable,
we shall be encouraged to approach, in an analogous spirit, some wider as-
pects of epistemology. The angel of topology and even more the demon of
algebra, in Bourbaki’s words, seem able to participate in a formalization
of macroscopic perception. Do similar perceptions correspond to similar
objects? What can be said about the composition of two perceptive pro-
cesses? This second question is perhaps the most interesting one and we are
inclined to think that macroscopic perceptive processes may be formalized
in algebraic terms. Examples are given by many physical or even biological
devices such as signal transmission (frequency filtering, temporal limitation
or sampling), and optical observation (spatial frequency filtering, framing).
It is possible to elaborate an algebra of what we call observation operators
[3,4].
This type of formalization clarifies many aspects of macroscopic per-
ception, not only deformation but also non-inversibility. For some reason
or another real perceptive processes have no inverse, it is not possible to
reconstruct the observed object from the knowledge of its perceived image
nor to separate two distinct objects having the same image. We have here
a limitation of epistemic nature, expressed formally by the non-inversibility,
in the mathematical sense, of the observation operator representing the pro-
cess. This observation operator, in the most simple cases, has to do with
the geometric concept of projection which reminds us of Plato’s cave [7].
According to this well known allegory, the shadows, observed at the rear of
the cave, are the only knowledge enchained men can acquire concerning stat-
ues moving between a bright fire and themselves. Not only do they obtain
thus a distorted, impoverished image of the statues but furthermore they
also attribute to the perceived shadows properties of the screen constituted
by the rear part of the cave, for example its bi-dimensionality, its irregu-
larities. This part of Plato’s metaphor gives a poetical expression of the
non-inversibility of perception and also of the subjective attribution, to the
observed universe, of structures inherent to the perceiving device, whether
purely physical or of biological nature. The use of non-inversible observa-
tion operators yields a mathematical formalization of this subjective inverse
transfer of structures [6,7] which has to do with the tendency to project
our inner world upon the external one, making it, even if abusively, more
familiar, endowing it with an appearance of pre-established harmony [9].
The above considerations on perception plead in favour of the pos-
ON THE POSSIBILITY OF A FORMALIZED EPISTEMOLOGY 105

sibility of a general formal epistemology [1,2] or even of a mathematical


epistemology, restricted to the macroscopic level [5].
In a less artificial way, it is possible to distinguish, for a physical or
biological structure, between an act of observation and decisional process.
In a formalized language, these steps can be represented by, respectively, an
observation operator and a decision operator, both of a same mathematical
nature. But, globally, we can define only one pragmatic operator, product
(to the left) of the observation operator by the decision operator, giving
access to an epistemology based upon decisions taken instead of perceptions
felt. Here again a formalization seems at hand, opening the way toward a
mathematical epistemo-praxiology [8,9].
So the cognitive aspect of epistemology seems at least partly formaliz-
able in algebraic terms, at the macroscopic level. This is the encouragement
we hoped to find. Is it possible to go further? A true formal epistemology
should not confine itself to the macroscopic world; it must also include quan-
tum aspects, thus encompassing both the macroscopic and the microscopic
levels. Could a general method be formed, able to help us find very basic
structures, more fundamental than those at hand. Let us try to proceed
along this way imperfectly delineated. We have, for example, an algebra
of macroscopic observation and also an algebra of microscopic observation.
There are between these structures at least some analogies which give some
hope to find a structure weaker than each of them, able to retain fundamen-
tal traits of both of them.
Other approaches seem possible. We can imagine an order, not nec-
essarily total, concerning natural systems. This hierarchy could be gen-
erated by the process of observation itself, from the coarsest to the most
delicate resolution level, from universe to particles. Epistemological analo-
gies of structures could be seen, being obviously more and more difficult
to cope with when approaching the microscopic level where they might be-
come meaningless, unless some fundamental formal aspect, to be discovered,
should still hold.
The construction of a formal epistemology, fundamental enough to
be relevant at all levels of knowledge, from the macroscopic to the micro-
scopic, is obviously a difficult task. The above considerations on a possible
mathematical epistemology valid at the macroscopic level, the indications
given about apparently plausible ways of research, are positive factors in
favour of the feasibility of a formal epistemology.

REFERENCES

1. M. Mugur-Schächter, “From quantum mechanics to universal struc-


106 R. VALLÉE

tures of conceptualization and feedback on quantum mechanics,”


Found. Phys. 23, 37-122 (1993).
2. M. Mugur-Schächter, “Une méthode de conceptualisation relativisée:
vers une épistémologie formelle apte à faire face aux complexités,”
Revue Internationale de Systémique 9, 963-994 (1995).
3. R. Vallée, “Sur deux classes d’‘opérateurs d’observation’,” C.R.A.S.
(Paris) 233, 1350-1351 (1951).
4. R. Vallée, “A note on algebra and macroscopic observation,” Informa-
tion and Control 1, 82-84 (1957).
5. R. Vallée, “Sur la formalisation mathématique en théorie de
l’observation,” 1ère partie in Actes du 7ème Congrès Internationale
de Cybernétique (Association Internationale de Cybernétique, Namur,
1974), pp. 225-232.
6. R. Vallée, “Observation, decision and structure transfers in systems
theory,” in Progress in Cybernetics and Systems Research, Vol. 1 (Sec-
ond European Meeting on Systems Research, Vienna, 1974), R. Trappl
and F. Pichler, eds. (Hemisphere Publishing, Washington, 1975), pp.
15-20.
7. R. Vallée, “La caverne de Platon revisitée,” in Praxis et Cognition
(Colloque International Praxis et Cognition, Cerisy, 1988), E. Bernard-
Weil and J. C. Tabary, eds. (L’Interdisciplinaire, Limonest, 1992).
8. R. Vallée, “Perception, decision and action,” Biological Systems 2, 43-
53 (1994).
9. R. Vallée, Cognition et Système, Essai d’Epistemo-Praxéologie
(L’Interdisciplinaire, Limonest, 1995).
Part Two
Constructive Contributions
7
QUANTUM MECHANICS VERSUS A METHOD OF
RELATIVIZED CONCEPTUALIZATION

Mioara Mugur-Schächter

Centre pour la Synthèse d’une


Épistémologie Formalisée (CeSEF)
47 Bd Georges Seurat
92200 Neuilly-sur-Seine, France

A general representation of the processes of conceptualization, founded upon


a descriptional mould drawn from fundamental quantum mechanics, is out-
lined. The approach is called the method of relativized conceptualization.
This stresses that the representation is not researched as a “neutral state-
ment of facts” but, from the start, as a method subject to definite descrip-
tional aims, namely an a priori exclusion of the emergence of false problems
or paradoxes as well as of any gliding into relativism. The method is charac-
terized by an explicit and systematic relativization of each descriptional step
to all the descriptional elements involved in this step, namely: the epistemic
action by which the object-entity is generated, the object-entity itself, and
the epistemic action by which the object-entity is qualified. Successive steps
which complexify progressively a given initial relative description, form an
unlimited chain of cells of conceptualization where the very first cell neces-
sarily is rooted in as yet strictly unconceptualized physical factuality, while
the subsequent cells consist of increasingly abstract descriptions that are
connected hierarchically. The chains interact at nodes where they branch,
thus generating an indefinitely evolving, complexifying web of relativized
conceptualization, free of ambiguities, and where each element stays under
control.
The method contains the posited assertion of a realism of which a
definite sort of minimality follows then inside the method. This generates a
clear distinction between illusory qualifications of “how-a-physical-entity-is-
in-itself” and models of this physical entity. Thereby a worked-out connection
with philosophical thinking is incorporated in the method. The method is

109
110 M. MUGUR-SCHÄCHTER

shown to entail a relativized genetic logic and a relativized genetic theory


of probabilities, more extended, respectively, than the classical logic and
the classical theory of probabilities. Both are rooted in physical factuality
whereby they merge in a unified representation of the logico-probabilistic
conceptualization.
The relations between the general method of relativized conceptual-
ization and the relativistic approaches in the sense of modern physics are
specified. These last ones, in contradistinction to the method exposed in this
work, are shown to concern exclusively the ways of constructing qualifiers of
object-entities so as to insure intersubjective consensus among correspond-
ing classes of observers, while the ways of generating the qualified object-
entities and the consequences entailed by these ways, are not considered:
Like in classical logic, as in all of classical thinking, the object-entities are
simply presupposed to always pre-exist available.
Traditionally, the emergence and elaboration of knowledge has always
been studied from a point of view founded on psychological and neurobiolog-
ical data and in the spirit of a neutral account of the “natural” phenomena;
the modern cognitivistic approaches continue this tradition. The approach
exposed in this work is probably the first one in which a systematic repre-
sentation of the processes of creation of knowledge is founded on strategic
data drawn from physics and, correlatively, is constructed from the start as
a method for the optimization of these processes themselves, accordingly to
definite aims.

Key words: quantum mechanics, method, descriptional relativities, concep-


tualization, epistemology.

“That a higher integration of science is needed is perhaps


best demonstrated by the observation that the basic enti-
ties of the intuitionistic mathematics are the physical ob-
jects, that the basic concept in the epistemological struc-
ture of physics is the concept of observation, and that psy-
chology is not yet ready for providing concepts and ideal-
isations of such precision as are expected in mathematics
or even physics. Thus this passing of responsibility from
mathematics to physics, and hence to the science of cogni-
tion ends nowhere. This state of affairs should be remedied
by a closer integration of the now separate disciplines.”—
E. P. Wigner [1].
QUANTUM MECHANICS VS. RELATIVIZED CONCEPTUALIZATION 111

Contents

1. INTRODUCTION
2. RETRO-PERSPECTIVE
2.1. Objectivity and Descriptional Relativities
2.2. “Existents”, or “Reality”, and Objectivity
2.3. The Polarity of Realism
2.4. Knowledge and Communicability
3. THE COGNITIVE SITUATION THAT LED TO THE
QUANTUM MECHANICAL DESCRIPTIONS

3.1. Historical Remarks


3.2. The Cognitive Situation Involved in Quantum
Mechanics and the Strategy Induced by it
3.3. Epistemological Universality
4. NORMS FOR DESCRIBING: THE METHOD OF
RELATIVIZED CONCEPTUALIZATION (MRC)

4.1. Preliminaries
4.2. The First Stage: a Presentation of MRC in Usual
Language
4.3. The Second Stage: an Ideographical Symbolization of MRC
4.4. The Third Stage: a Scheme of a Mathematical Representation
of MRC in Terms of the Theory of Categories
5. ILLUSTRATIONS OF THE FUNCTIONING OF MRC
5.1. Classical Logic versus the MRC-logic of Relative
Classes of Cognitive Actions
5.2. MRC versus Probabilities
6. MRC VERSUS OBJECTIVITY IN THE SENSE OF THE RELATIVISTIC
APPROACHES FROM MODERN PHYSICS
7. FINAL REMARKS

1. INTRODUCTION

This work is submitted here as an illustration of how a formalized epis-


temology can be researched accordingly to the principles expressed at the
beginning of this volume. Indeed, what I call the method of relativized con-
ceptualization can already be regarded, I think, as a first but rather firm
construct on the way toward a fully satisfactory formalized epistemology.
Though initially induced by—specifically—the cognitive strategy brought
forth by the analysis of fundamental quantum mechanics, this construct, by
the universality of the epistemological essence drawn therefrom and by the
112 M. MUGUR-SCHÄCHTER

way of elaborating it, possesses an unrestricted relevance concerning any


process of conceptualization.
In a certain sense, the way in which the method of relativized con-
ceptualization is offered here is highly artificial. This method developed in
my mind very slowly, while periodically, year after year, in the course of
my University lectures on elementary and advanced quantum mechanics, on
probabilities and on information theory, I was once more scrutinizing the
formalisms of these three theories. This recurrence, by a process of integra-
tion, produced the method of relativized conceptualization at the same time
with what I now call meta[quantum mechanics]1 and which—a posteriori -
appears as a major illustration of the method, belonging organically to it.
But meta[quantum mechanics] is far too technical to fit into this volume.
So I chopped it off and healed the scars by a brief informal preliminary
exposition of—strictly—only the essence of the considerations on quantum
mechanics which triggered the method of relativized conceptualization. The
result might appear somewhat strange due to restricted access to the struc-
ture of mathematical features which determined it from inside fundamental
quantum mechanics and which, together with the emerging method itself,
guided the modified reconstruction of quantum mechanics which I call meta
quantum mechanics, which in its turn illuminates the method. But, on the
other hand, any method, once constructed, should be able to convince by
itself. So, insofar as a method, such as the one exposed here, fails to do that,
it simply is devoid of a genuine inherent conceptual and operational value.
Inside the community of physicists, this work will appear as exte-
rior to all the present-day main streams. Of course, there have been many
famous physicists who have tried to understand how quantum mechanics
works, what it really asserts, and what it leaves open. But, as far as I know
at least, no physicist as yet has tried to work out explicitly, specifically, and
systematically, the universal epistemological implications of the quantum
mechanical formalism. The novelty of this aim imprints a peculiar charac-
ter upon the approach. This is why physicists might feel disconnected while
reading what follows. In order to nevertheless gain their attention and fix
it upon the epistemological problems dealt with in this work, I take the
liberty to claim once more that, inside meta[quantum mechanics], the re-
sults established in this work entail a clear optimization of the formalism
of fundamental quantum mechanics—with respect to its own descriptional
aims—and that they furthermore yield a thorough intelligibility of this for-
malism, which cannot but enhance the efficiency in dealing with the basic

1
Partial indications on meta[quantum mechanics] can be found in various other works
(Refs. 11 to 17). A complete final account is not yet available but I hope will be published
soon.
QUANTUM MECHANICS VS. RELATIVIZED CONCEPTUALIZATION 113

problems of modern physics in general.


The philosophers, with respect to their own knowledge and criteria,
will certainly find insufficiencies in this work. I apologize to them in advance:
With the means available to me, I have tried to build a solid bridge between
physicists and philosophers. Others might want to improve on it in various
ways.
Of course, a formalized epistemology, in the full sense assigned to
this term in the introduction to the present volume and in the contributions
from the first part, should incorporate methodological procedures explicated
also from other modern disciplines besides quantum mechanics, in particular
from mathematics, informatics, biology, cognitive and neurological sciences,
linguistics, and philosophy. Some steps in this direction can be found in
other contributions to this volume (cf., in this volume, the contributions
of Robert Vallée, Élie Bernard-Weil, Giuseppe Longo, Evelyne Andreewsky,
and Vincent Schächter).

2. RETRO-PERSPECTIVE

Before entering upon the exposition of the method of relativized conceptu-


alization, I shall briefly sketch out in what historical retro-perspective it fits
in.

2.1. Objectivity and Descriptional Relativities


The concept of scientific objectivity is undergoing a revolution. The classi-
cal concept of objectivity was tied with the posit that science just discov-
ers truths that are independent of any human aim-and-action, pre-existing
“out there” such as they appear when discovered. But throughout the last
century this view kept receding. It became increasingly clear that objec-
tivity in the classical sense was an illusion; that scientific knowledge is con-
structed under certain constraints which characterize the epistemic situation
and the epistemic aim of the acting observer-conceptor and imprint upon
the result non-removable descriptional relativities to this situation and this
aim. More or less implicitly, awareness of quite essentially involved [(epis-
temic situation)-(epistemic aim)] structures, developed steadily, perturbing
the classical conception about objectivity while instating a new concept of
objectivity in the sense of inter-subjective consensus.
So far however, only few have already gained an explicit and clear
awareness of this evolution. Correlatively, on a metalevel, a fully organized
and general view on the epistemic actions by which scientific inter-subjective
consensuses are achieved is still lacking. What, exactly, in scientific consen-
sus, insures subjection to also what is called reality and truth, thereby tran-
114 M. MUGUR-SCHÄCHTER

scending mere conventionality and withstanding relativism? How, in what


a sense and to what a degree, is reference insured? How, exactly, do the
involved human aims and features come into play? What particular sorts
of strategies are put to work in order to construct scientific inter-subjective
consensuses? While such questions struggle for definite answers, the inertial
forces that work inside language bring forth again and again the same old
word—objectivity—to designate indistinctly either the emerging new con-
cept, or the classical one. This favours the persistence of many circularities
and confusions.
Let us now consider physics. The employed cognitive strategy varies
radically as one shifts from fundamental quantum mechanics to the theory
of relativity and to relativistic approaches in general.
Fundamental quantum mechanics incorporates—implicitly—a pecu-
liar type of “basic” descriptional relativities which insert the very first stra-
tum of conceptualization, deep into purely factual physical reality. The de-
scriptional relativities of this basic type, when entirely explicated and then
generalized, lead toward a recasting of epistemology. The main lines of this
major consequence of the quantum mechanical strategy for constructing
knowledge are captured in the method of relativized conceptualization. This
method, while it strongly connects modern physics with philosophy, will be
shown to entail also a non classical unification between logic, probabilities,
and set-theory.
On the other hand, inside the theory of relativity and more generally
inside the whole class of relativistic approaches, another sort of methods for
constructing inter-subjective consensuses have been developed. These, much
better recognized than those involved by fundamental quantum mechan-
ics, are only very indirectly and loosely connected with physical factuality.
They are quasi exclusively dominated by abstract constraints of a logico-
mathematical nature imposed upon the representational features tied with
“states of observation”. The formal constructs entailed by this sort of con-
straints manifest a vertiginous growing of the degree of conceptual freedom
displayed by modern physicists in the representations of physical reality. In
these constructs one can again identify forms of the general tendency, in
modern physics, to merge with epistemology and philosophy.
So in modern physics, objectivity, quite generally, means constructed
inter-subjective consensus founded on descriptional relativizations that point
toward an underlying trend of unification of physics, with epistemology and
philosophy.
The method of relativized conceptualization, which is the core of this
work, was crystallised out of this trend.
QUANTUM MECHANICS VS. RELATIVIZED CONCEPTUALIZATION 115

2.2. “Existents” or “Reality”, and Objectivity


The existence, for each human being, of an inner psychical reality, prob-
ably has never been doubted by any normal person. Following Descartes,
Berkeley, Kant, Husserl, the philosophers place it explicitly at the bottom
of any knowledge. Physicists have never denied it. Nor did common sense.
And nevertheless, paradoxically, for most people the quintessence of what is
called reality, of what is hold to be “genuinely” existent, is the exterior and
physical reality; even if this or that marginal individual happens to perceive
the exterior physical reality as less certain than his own inner reality, or
even—at the solipsistic limit—as wholly illusive.
This entangled hierarchy has multiple manifestations. For instance, it
is striking that concepts, and more generally knowledge, languages, science,
are seldom explicitly taken into account as constituents of reality, strictly
speaking. It is true that Teilhard de Chardin did so (this is his major speci-
ficity); that Karl Popper [2] asserted “three worlds”, the physical reality,
the states of consciousness, and knowledge, arts, cultural facts; and that,
no doubt, other important examples can be found. But, on the other hand,
up to this day the debate on the existents (do the unicorns exist?) still
continues among logicians [3]. Platonism has enemies as much as adepts,
etc. And, more or less implicitly, a general tendency can be observed to set
aside the word reality for designating exclusively what is posited to exist
outside any psychism and moreover is physical. A larval form of this ten-
dency is present in particular in the reductionist view according to which
anything which at a first sight seems not to consist of exclusively physical
entities, in fact is strictly deducible—without any loss—from the existence
and laws of the physical reality alone. This view, favoured by a loose con-
tact between philosophers and scientists, is still quite active in many eminent
minds, notwithstanding that most philosophers perceived it as naı̈ve and il-
lusive already since Descartes, while since Kant they almost unanimously
banished it explicitly and radically.
On the other hand, Einstein relativity and then—otherwise—
quantum mechanics, induced a stream of change into the content assigned
in physics to what is called truth and objectivity. The main contribution
to this stream consists of deliberate constructions of symmetries concerning
the processes of qualification of the considered object-entities, symmetries
tied with groups of operations of transformation of the state of observation.
But furthermore other modern developments of the “exact” thinking, log-
ical, mathematical, informatical, also contribute to this stream, by direct
elaboration of grammars (syntaxes) admitting of models (interpretations),
by algorithms for reconstructing phenomena by simulation instead of repre-
senting them by assertions and proofs, etc. Now, all these new approaches
116 M. MUGUR-SCHÄCHTER

are methods for constructing inter-subjective consensus concerning results


of manners of conducting descriptional actions in order to reach a definite
aim of knowledge. They all involve an explicit teleological dimension where
factors of various natures—psychical or biological or physical, factual or
abstract-conceptual—co-operate inside an organic whole. This amounts to
an implicit deletion of the classical belief that consensus manifests a pre-
existing objective truth which has to be just learned, apprehended.
This evolution induces the scientific thinkers into rediscovering by
themselves certain basic features of Kant’s constructivist view on objectivity
[4,5]. This, among those who work in the foundations of science, generates an
increased receptivity with respect to the philosophical thinking sedimented
since millennia. While on the other hand the philosophers tend more and
more to concentrate upon the methods and languages that emerge inside the
sciences, trying to bring forth the new philosophical implications of these.
Globally, philosophy and the sciences are meeting in a process of
re-elaboration of the concepts of reality and objectivity.
I shall now go to the bottom of this process, but specifically from
the point of view of a physicist. I shall focus upon the content of the very
first layer of the emergence of the inter-subjectively known, such as it can be
characterized when the involved biological processes, though fully recognized
to play a key role, are not themselves the object of investigation (as in the
modern researches on cognition and consciousness [6,7,8]) but are regarded
as only a datum to be explicitly taken into account.

2.3. The Polarity of Realism2


Kant stated explicitly that exclusively phenomenal appearances are known
in a non-mediated way. The word phenomenon designates here a conscious
event from an individual mind, already cast in the a priori forms of human
intuition, time and space. This conscious event can be conceived by the man
who experiences it as reflecting, or not, some object-entity; but in any case it
somehow bears the mark of the acting human body-and-mind structure, in a
non removable and inextricable way. This is the foundation of the well-known
Kantian postulate of impossibility to know reality such-as-it-is-in-itself, i.e.,
independently of any structure interposed by the observer-conceptor.
It is curious to note that this famous Kantian impossibility concerns
exclusively the reality that is exterior to the mind. Indeed, if one chooses
to point via this same term, reality, toward any sort of existent, no matter
whether assigned to the exterior universe or to some interior universe, this
rather natural extension of language generates an exception to Kant’s pos-

2
This section has benefited from precious remarks made by Hervé Barreau.
QUANTUM MECHANICS VS. RELATIVIZED CONCEPTUALIZATION 117

tulate, a huge one. For on the one hand this extension of language entails
that also a phenomenon from an individual mind is an element of reality.
But on the other hand a phenomenon, by definition, is just that what ap-
pears to the mind where it emerges. So, for the sake of self-consistency, a
phenomenon, as such, has to be posited to be known by the mind where it
emerges precisely such-as-it-is-in-itself. To assert the contrary would simply
be a logical contradiction in the construction of the whole consisting of [lan-
guage and what it is posited to refer to]. Later the considered phenomenon
might be perceived differently by the person who experienced it, or if it is
communicated to another mind its description might there be variously in-
terpreted, in psychoanalytical terms, or biological ones, etc. But in all such
cases one is in fact speaking of another (meta)object-entity that is related
with the initial phenomenon but is not identifiable with it. And this new
(meta)object-entity, in its turn, again must be posited to be known by the
mind where it emerges, such-as-it-is-in-itself, etc. (This same point has been
made also by other authors, e.g., Goodman [9].) This characteristic of the
inner phenomena, however, is not in the least a “problem”. On the contrary
it seems to be in deep harmony with the Cartesian cut.
Indeed, the fact that an entity from an inner individual universe has
to be considered to be precisely such as it is perceived, can be considered
to mark a polarity of reality with respect to knowledge, by which, while the
exterior reality never can be known such-as-it-is-in-itself, any piece of interior
reality—at the time when it emerges in this or that individual mind—can
only be known by that mind such-as-it-is-in-itself, whereby its “truth” is
beyond any doubt (or is a qualification devoid of pertinence, which amounts
to the same thing), so it is endowed with the Cartesian sort of pre-eminence.

2.4. Knowledge and Communicability


But let us come back to the fact that a phenomenon, by definition, can
only exist inside an individual mind. At the time when a given phenomenon
emerges in an individual mind, it is known there without being also com-
municated. The subject can even know it without having ex-pressed it for
himself: it can remain an unexpressed, a-symbolic individual psychical fact,
chained to, and somehow melted to a certain degree in the interior uni-
verse where it happened. On the other hand, according to thinkers who
know Kant’s work deeply, in the Kantian view any scientific objectivity is
constructed by a method of “legalization” of the primary phenomenal ap-
pearances. In this respect, Jean Petitot ( [4]) writes:
“The object of experiment, of scientific knowledge, is not given in
the donation of the phenomenon. It emerges by objectual legalization
of phenomena. So, apart from a descriptive dimension, any scientific
118 M. MUGUR-SCHÄCHTER

knowledge presupposes in its very principle also a prescriptive, a nor-


mative dimension, that is constitutive of objectivity . . . . In Kant’s
work—so concerning classical mechanics—the method consists essen-
tially in interpreting the categories of objectivity 3 by starting from
the instances of donation of the phenomena, that is, by starting from
the forms of phenomenal manifestation. Since the interpretation of
the categories of objectivity is operational only if it is mathematical,
the forms of phenomenal manifestation themselves must be mathe-
matized.”
But such a legalization involves communicability. So, how is the trans-
position of a phenomenon into communicable symbolizations to be set up?
Here, at precisely this point, one is confronted with an obscure zone
where is located—undefined—the structure of the very first stage of inter-
subjective conceptualization, that on which the whole subsequent inter-
subjective conceptualization is founded, so also objectivity in general and in
particular scientific objectivity. Kant did not deal with this question.4 And
as far as I know, up to now the philosophical thinking did not yet concen-
trate constructive efforts upon this zone. But it produced already important
“negative” developments. The whole question of reference on which Quine
[10] and Putnam [11] for instance achieved so deep and compelling analyzes
in order to establish the frontiers of the domain inside which language con-
fines knowledge, takes its sources precisely in the above mentioned obscure
zone.
Now, in so far that one agrees that any transposition of a phe-
nomenon, in communicable terms, amounts to a description, the content
of this obscure zone can be more narrowly pointed toward by the following
formulation:
Nothing else but descriptions can be known in an inter-subjective way,
neither exterior factual entities “themselves”, nor non-described phe-
nomena.
This specification is far from being trivial: it focuses the attention
upon the primary importance of the emergence of communicability. Com-
municability in general as a larger basis for the particular sort of commu-
nicability that is normed scientifically. By way of consequence it establishes
3
The “dynamical” (physical) categories of substance, of causality and of interaction,
the categories of quality and quantity, and the “modal” categories of possibility (poten-
tiality, virtuality), of reality (actuality) and of necessity.
4
As Hervé Barreau puts it, in Kant’s view the phenomena seem to emerge directly
Newtonian, already cast in scientific Euclidean spacetime. Any concern about geneses of
the type of those examined later by Husserl, Bergson, Piaget, and so many others, is
absent in the Kantian work.
QUANTUM MECHANICS VS. RELATIVIZED CONCEPTUALIZATION 119

the interest of defining a canonical structure for what is called a description,


a normed form of the descriptions, a mould into which to pour in an agreed
way any transposition of a phenomenal appearance, in communicable terms.
It establishes the inadequacy of a notion of—directly—a “scientific legaliza-
tion of phenomena” which omits, hides into the non-analyzed and non legal-
ized, the more basic stage of accomplishment of descriptions. Indeed, only a
conveniently structured general norm for accomplishing descriptions could
act as a universal inter-subjective reference permitting to gauge against it
any procedure for describing, the natural descriptional procedures, as well
as, in particular, the various procedures for a “scientific” legalization of the
descriptions of object-entities of any kind, so also of conscious phenomena.
These procedures could then be all qualified, compared, understood, inside
a common frame where a certain unity is set in advance beneath the speci-
ficities tied to this or that descriptional approach.
But how, according to which criteria, shall we identify the canonical
form to be required for any description?
It is quite remarkable that the answer to a question of this nature and
of such generality can be drawn from a physical theory. For it is quantum
mechanics which shows the way, if the descriptional aim chosen in it and
the strategy practised in order to reach this aim, are thoroughly explicated.

3. THE COGNITIVE SITUATION THAT LED TO THE


QUANTUM-MECHANICAL DESCRIPTIONS

3.1. Historical Remarks

A cognitive situation like that one involved in the quantum mechanical for-
malism, so extreme, had never been dwelt with systematically before the
construction of quantum mechanics. A cognitive attitude like that one in-
duced by the mentioned cognitive situation, so radically creative, had never
before been organized. But when a theory of “microstates” started being
researched, the involved cognitive situation acted, without getting explicit
for that. The various well-known contributions from Plank, Einstein, Bohr,
de Broglie, Schrödinger, Heisenberg, Born, Pauli, von Neumann, Dirac, etc.,
led to a coherent whole because they all had to satisfy, more or less implic-
itly, the same strong and peculiar constraints, those imposed by the involved
cognitive situation. (But, and it is curious to find this out, none among the
so numerous and eminent contributors did fully grasp the new epistemologi-
cal essence of the emerging construction). There has been no equivalent, for
quantum mechanics, of a Newton, a Maxwell, a Carnot, a Boltzmann, or an
Einstein.
120 M. MUGUR-SCHÄCHTER

The construction of the quantum mechanical formalism has been or-


chestrated by an impersonal, very peculiar cognitive situation.

This might explain why the formalism, notwithstanding its remark-


able efficiency, is up to this very day thought to possess a cryptic character
and to involve problems. These problems however, over and over again, are
much more referred to the formalism itself than to the cognitive situation
which commanded the form of the algorithms. While, as far as I know at
least, the cognitive situation has never been explicitly and thoroughly re-
considered for itself. So, hidden beneath increasingly complex formal devel-
opments and surreptitious mutations of the theory as a whole, its seminal
epistemological implications could remain for ever devoid of contour, their
substance anonymously absorbed and assimilated in the process of the evolu-
tion of physics. This would be a big loss. Only what is named and described
explicitly gets contour and can act sharply and deeply.
In what follows, I withstand this decay. In a very synthetic and simple
way I shall outline the main specific epistemological features of the cogni-
tive situation involved in the quantum mechanical formalism. Thereby, in
fact, I achieve a first step in the direction of what I call meta[quantum me-
chanics]. Indeed, as already mentioned, this re-formalization of fundamental
quantum mechanics, of which certain rather elaborate and much more tech-
nical elements (but never the whole so far) have been exposed in other
works [12,13,14], is founded on—both—the basic considerations exposed in
Sec. 3.2 and on the fully elaborated method of relativized conceptualization.
Here, however, the aim is exclusively to bring into evidence the source of
the method. The very simple exposition that follows (which summarizes two
earlier non-specialized presentations ([15,16]) should suffice.

3.2. The Cognitive Situation Involved in Quantum


Mechanics and the Strategy Induced by It
A description involves a definite object-entity (object-of-description) and
qualifications of it. The basic object-entities of quantum mechanics are what
is called states of microsystems (microstates).5 These are hypothetical enti-
ties that no human being (in the present-day sense) will ever perceive. The
5
The stable microsystems themselves (electrons, protons, neutrons, etc.) have first
been studied in atomic and nuclear physics where they have been characterized by specific
“particle”-constants (mass, charge, magnetic moment). Changes of stable microsystems
(creation or annihilation) are studied in nuclear physics and in field-theory. States of stable
microsystems—microstates—are specifically studied in fundamental quantum mechanics
where they are characterized by probabilistic distributions of values of state-“observables”
(for Dirac the word “sate” is short for “way of mouvement” of a dynamical system (mi-
crosystem)).
QUANTUM MECHANICS VS. RELATIVIZED CONCEPTUALIZATION 121

obtention for them of qualifications endowed with some sort of stability,


raises difficult and deep questions. Nevertheless quantum mechanics exhibits
a very performing description of microstates. This manifests a descriptional
strategy that has succeeded to overcome the epistemological difficulties. We
want to explicate this descriptional strategy.
Let us consider first the basic object-entities of the quantum mechan-
ical descriptions, microstates. Since they cannot be perceived, such object-
entities cannot be made available for study by just selecting them inside some
ensemble of pre-existing entities. Nor can one study them by just examining
observable marks spontaneously produced on macroscopic devices by admit-
tedly pre-existing natural microstates: no criteria would then exist for de-
ciding which mark is to be assigned to which microstate. The unique general
solution, then, is to first accomplish a known and repeatable macroscopic
operation posited to generate a given though unknown microstate, and to
try afterward to somehow manage to “know” the generated microstate.
Consider the hypothetical microstate produced by a given opera-
tion of state-generation. The plan is to acquire concerning it informations
cast in certain pre-established terms, involving what is called “position”, or
“momentum”, or “energy”, etc. The grids for the desired sorts of qualifica-
tion are conceived beforehand, quite independently of the generated object-
microstate, and with respect to these grids the object-microstate emerges
in general still entirely unknown, still strictly non-qualified. This assertion
is not in the least weakened by the fact that the presuppositions of the ex-
istence of microstates and of the emergence of a given sort of microstate
when a given operation of state-generation is realized, insert already the
generated microstate into a net of pre-conceptualization, so of a kind of pre-
posited knowledge: the generated microstate emerges non-perceptible, so a
fortiori still entirely non-singularized from the specific points of view ex-
pressed by the definitions of the grids of its desired further qualifications.
But on the other hand it emerges also relative, in a non removable way, to
the employed operation of state-generation, and this permits to label it: it is
a result of this—known—macroscopic operation of state-generation. Let us
immediately embody this possibility. Let us symbolize by G the considered
operation of state-generation and by msG the corresponding generated mi-
crostate. Though in this incipient stage the symbols G and msG are devoid
of any mathematical representation, their introduction is very important.
Indeed it instates inside the realm of the communicable, the fact that the
generated microstate, though unknown, is nevertheless captured, in the pe-
culiar sense that one can now produce as many copies of it as necessary
and subject each copy to some subsequent operation of examination, while
communicating clearly what one does, by words and signs. This amounts to
122 M. MUGUR-SCHÄCHTER

having achieved a sort of a-conceptual definition of an infinite set of repli-


cas of the object-entity called a microstate generated by G and symbolized
msG . A purely factual and nevertheless communicable definition. This is
very remarkable because it circumvents the lack, for defining msG , of any
predicate: G is not a qualification of msG , it is the way of producing it.
Thereby one of the extremities of the chain of information that was
to be started, is now fixed.
Once the first stage, of production of a “given” object-entity, has
thus been achieved, one can enter upon the second stage, of construction
of a certain new knowledge concerning specifically the generated object-
entity. Now, the object-entity denoted msG , such as it emerges from the
operation G that generates it, in general does not reach the level of what
is observable by man. So it has now to be brought to trigger on this level
some observable manifestations. Furthermore these manifestations have to
be endowed with significance, namely with precisely the researched kind
of qualifying significance. In order to reach this new aim, measurement in-
teractions M (X) with macroscopic measurement devices are organized for
measuring the quantum mechanical dynamical quantities X; X runs over
the set of dynamical quantities—position, momentum, energy, etc.—that
are mathematically defined inside quantum mechanics and M (X) desig-
nates the process by which X is measured. The formal representations of
the measurement-interactions M (X) are mainly conceived in a peculiar sort
of prolongation of the classical mechanics. Thereby, implicitly, history and
models come in ( [15], [16]). The practical realizations of the measurement
interactions M(X) are planned such as to produce a perceptible set of marks
{µX } upon a convenient X-registration-device of an apparatus A(X) “good”
for measuring X on microstates. What this means is quite non-trivial. In fact
the processes M (X) are produced by what is called the apparatus A(X).
Each set {µX } of observable marks, once realized, is interpreted, it is coded
in terms of a value Xj of the quantum mechanical dynamical quantity X
(Xj is called an eigenvalue of X); j is a discrete or continuous index. Which
j corresponds to which sort of mark has to be specified so as to define a sta-
ble code-language. The coding-rules are determined by the formal quantum
mechanical definition of X and by the specification of the interaction chosen
as a measurement process M (X).
Codability in this sense—a rather complex operation—is a central
condition for M (X) to be acceptable as a “measurement” process of
X, so for A(X) to be acceptable as a “good” apparatus for measuring
X.
In this manner—by a complex interplay of inherited pre-
conceptualizations, of assumptions, implicit models, macroscopic operations,
QUANTUM MECHANICS VS. RELATIVIZED CONCEPTUALIZATION 123

theoretical representations, and of calculi and codings—are achieved the ba-


sic quantum mechanical qualifications of microstates.
Of microstates, indeed? Let us avoid inertial steps in the way
of speaking, and check the pertinence of each verbal expression. For it
seems clear that in general a measurement interaction must be imagined
to change the microstate initially created by the employed operation of
state-generation, possibly quite radically in certain cases; so the observ-
able marks emerge indelibly relative to the employed measurement process.
Which means that these marks characterize globally the measurement in-
teraction, not separately the supposed object-microstate. One can however
cling to the fact that the observable marks are relative to also the initially
created microstate, while the type of change undergone by this microstate
during a measurement interaction is ruled in an admittedly known way by
what is called a measurement process M (X). One has then to take fur-
thermore into account that two distinct processes of change of the initially
produced object-microstate, corresponding to two distinct measurement in-
teractions M (X) and M (X 0 ) of two different quantum mechanical dynamical
quantities X and X 0 6= X, in general cover two different spacetime domains.
When this happens, the measurement-processes M (X) and M (X 0 ) cannot
be both simultaneously achieved starting from one single replica of a mi-
crostate msG : in this sense these two measurement interactions are mutually
incompatible. So, if one wants to obtain observable qualifications involving
the microstate msG , in terms of eigenvalues Xj and Xk0 of both X and X 0 ,
one has in general to generate more than only one replica of msG because
one has to achieve two sorts of successions [(a given operation G of state gen-
eration ),(a measurement process M (X) on the supposed result msG of G)]
(in short successions [G.M (X)]), namely [G.M (X)]) and [G.M (X 0 )]), (the
chronometer being re-set at the same initial time-value before the realiza-
tion of each pair). Furthermore even the measurement on a microstate msG ,
of only one quantum mechanical dynamical quantity X, when repeated via
the corresponding succession [G.M (X)], in general does not yield systemat-
ically one same eigenvalue Xj ; in general the results are distributed over a
whole spectrum {Xj, j ∈ J] of possible eigenvalues of X (J: an index set,
discrete or continuous). Moreover a given eigenvalue Xj can in general be
obtained also with other microstates msG0 6= msG corresponding to other
operations of state-generation G0 6= G. In short, a stable information—if it
can be obtained—cannot concern isolately one individual microstate msG .
It necessarily concerns some pair [G.M (X)] in which the measurement inter-
action M(X) is also involved. And, furthermore, in general a pair [G.M (X)]
has to be repeated in order for us to become able to assert a stable result. So
a whole—big—set of replicas of the microstate generated by G is involved.
124 M. MUGUR-SCHÄCHTER

This means that the observational invariants that can be obtained by the
help of pairs [G.M (X)] can only consist of probability laws p(G, X) defined
on the spectra {Xj } of the quantum mechanical observables X. Now, noth-
ing insures a priori the existence of such probability laws. This existence is
not a logical necessity. And if no probability laws associated with the vari-
ous pairs [G.M (X)] were found, one would be obliged to finally give up the
aim to construct some stable observable knowledge concerning microstates.
But in fact it turns out that probability laws p(G, X) do arise, for each
pair [G.M (X)]. So, by a very big number of repetitions of pairs [G.M (X)]
where X runs over the set of all the dynamical quantities defined inside
quantum mechanics, classes {Xj , j ∈ J} of eigenvalues are obtained, cod-
ing for sets of registered marks that are mutually incompatible in the sense
specified above, and over these probability laws p(G, X) are found. These
probability laws, like also the concerned observable events Xj and their in-
dividual probabilities p(G, Xj ), are relative to both the involved operation
G of state-generation and the involved dynamical quantity X 6 .
But thereby knowledge of the studied object-entity itself, the hy-
pothetical microstate labelled msG , remains non-extracted from inside the
observable results of the pairs of operations [G.M (X)]. The descriptional
strategy imposed by the cognitive situation leads to observable qualifica-
tions that can be posited to involve this object-entity, but cannot be as-
signed to it alone, separately from the macroscopic operations G and M (X).
This is a serious hindrance when one wants to think and speak about—
specifically—“microstates”. To overcome this handicap one can make use of
a sort of an ad hoc conceptual construct. Instead of speaking of the probabil-
ity p(G, Xj ) of this or that observable event Xj tied with a pair [G.M (X)],
one can, equivalently, speak of the potentiality of the microstate msG itself
to produce with probability p(G, Xj ) the observable manifestation Xj if a
measurement M (X) is performed on this microstate msG . Which centres the
thought-and-locution upon the microstate msG itself. In this way the concept
of relative potentialities of observable manifestations permits to found upon
the observable marks µX obtained by measurement interactions M (X), a

6
The fact that repetitions of pairs [G.X] are necessarily involved in the construction of
an observable knowledge concerning the hypothetical microstates, entails quite non-trivial
conceptual questions. These, because no specific language for dealing with them conve-
niently has been constructed, have led to what is called the “problem of the completeness
of quantum mechanics” [15, 16]. Here we slip over these questions because inside the
method of relativized conceptualization we shall deal in detail with the sources of a gen-
eralalized equivalent of this problem. Let us only note that one probability law p(G, Xj )
is not considered to be sufficient for an unambiguous characterization of the involved mi-
crosystem msG ; to achieve such a characterization it is necessary to exhibit at least two
such probability laws corresponding to two mutually incompatible observables X and X 0 .
QUANTUM MECHANICS VS. RELATIVIZED CONCEPTUALIZATION 125

standard way of speaking about the microstate msG itself, namely in terms
of potential and relative hypothetical “properties” which are “possessed”
by it alone, before the changes undergone during the measurement interac-
tions that led to observable marks µX characterizing these interactions as a
whole. But, mind that, what is achieved in this way is not more than just
a model that should by no means be confused for an impossible specifica-
tion of how-msG -really-is-in-itself. A very remote and poor, minimal sort of
model, in fact, because of the non removable double relativization, to G and
to M (X), and of the only retroactive, hypothetical, potential and relative
character of the assigned potential “properties”. But nevertheless a model
that introduces a standard way of speaking of the posited microstate itself.
Which is a precious alleviation for thinking of it.
The spacetime incompatibilities between different measurement inter-
actions M (X) achieved on distinct replicas of the microstate msG generated
by a given operation of state-generation G, entail, in terms of the minimal
model specified above, that:

The set of all the physical processes of actualization of the various


relative potentialities of observable manifestations Xj assigned to a
micro-state msG generated by a given operation of state-generation
G, falls apart into a set of mutually incompatible classes of actualiza-
tion. This brings forth a probabilistic whole of a new type, with a tree-
like spacetime structure, and involving triadic chains with potential-
actualization-actualized links.

I called this structure the quantum mechanical probability tree of the


operation of generation G [12-14]. By systematic reference to the quantum
mechanical probability trees, the quantum mechanical formalism can be un-
derstood clearly and in full detail. This sort of reference constitutes the
key-procedure for the construction of what I call meta[quantum mechanics]
(note 1).
The preceding account, brief and simple as it is, contains, I think,
the whole essence of the quantum mechanical descriptional strategy and of
the type of results brought forth by it. It shows clearly how epistemic aims,
physical epistemic operations (required for acquiring the desired knowledge),
hypotheses concerning the contents or the results of the epistemic opera-
tions, observable facts, codings of these in terms which assign meaning to
them, and more or less explicit modelizations which come constantly in, are
brought to form together coherent wholes called “states of microsystems”.
What sort of objectivity do such descriptions insure? The knowl-
edge constructed by the quantum mechanical descriptions is endowed with
objectivity in the following sense. All the physicists who, working at dif-
126 M. MUGUR-SCHÄCHTER

ferent spacetime locations, are in states of observation constantly devoid of


acceleration with respect to one another, obtain the same probabilistic dis-
tributions p(G, X) if they apply the quantum mechanical prescriptions for
obtaining observable results concerning a given pair [G.M (X)] as well as the
galilean transformations of spacetime coordinates when a passage from an
inertial referential to another one is involved. This means that the quantum
mechanical probability distributions p(G, X) are invariant with respect to
Newtonian changes of the spacetime coordinates, they are physical “New-
tonian laws” associated with the considered pairs [G.M (X)]. That is, they
are pieces of inter-subjective consensus involving physical operations and
facts, insured inside a particular but “sufficiently” large class of different
observer-conceptors.
3.3. Epistemological Universality
It appeared above that the quantum mechanical descriptions are the re-
sult of a deliberate construction of communicable knowledge, a construction
founded on the systematic relativization to pairs of operations [G.M (X)].
In order to achieve a quantum mechanical description of a microstate it has
been necessary:

(a) to achieve the epistemic action denoted G that introduces the object-
entity, independently (in general) of any epistemic action by which
this object-entity could be qualified;
(b) to achieve the epistemic actions that lead to qualifications of the
object-entity;
(c) to realize both these distinct sorts of epistemic actions in a radically
creative way, by first generating— physically, in spacetime—an object-
entity that did not pre-exist, instead of just selecting it among already
available physical objects, and by then generating, again physically,
in spacetime, also observable manifestations of the previously gen-
erated object-entity, instead of just detecting pre-existing properties
possessed by this entity;
(d) to realize a big number of replicas of the pair [G.M (X)] for each quan-
tum mechanical dynamical quantity X, in order to construct invariant
probabilistic qualifications (because in general no individual invariants
are found).

Now, this is a maximally displayed and creative way of achieving de-


scriptions, where all the involved relativities are active. It is crucial to realize
clearly that such a degree of display and creativity is absent in most of our
current classical conceptualizations such as they are reflected by the nat-
ural languages as well as by logic, probabilities, physical theories, Einstein
QUANTUM MECHANICS VS. RELATIVIZED CONCEPTUALIZATION 127

relativity included. In the classical conceptualizations it has always been pos-


sible to suppose more or less implicitly that the considered object-entities
pre-exist to the descriptional process, that they are “defined” in advance by
properties which they possess already actualized and independently of any
act of examination. As long as the peculiar aim of describing states of mi-
crosytems had not yet been conceived, this supposition never led to noticed
difficulties. Therefore, classically, a description is conceived to consist ex-
clusively in the detection of one or more among the actual properties of the
pre-existing object-entity. The question of how the object-entity is introduced
is entirely skipped. As for the dynamical evolution that creates knowledge
of a qualification, it is shrinked into one static act of mere detection. With
respect to the quantum mechanical descriptional scheme, this last classical
contraction is the source of the nowadays most explicitly known differences
between quantum logic and probabilities, and classical logic and probabili-
ties. While in fact the—ignored—consequences of the explicit consideration
of the way in which the object-entity is generated, are still much deeper.
It is however noteworthy that, while in classical logic and classical
probabilities—the two most fundamental classical syntactical structures—
the quantum mechanical descriptional scheme is not apparent, this scheme
nevertheless is explicitly involved in many classical and quite current epis-
temic situations and procedures. Indeed, once one has clearly perceived the
peculiar and very difficult epistemic situation dealt with in quantum me-
chanics, as well as the descriptional strategy that permitted to dominate it,
a very paradoxical inversion arises, by a sudden variation that reminds of
those which make appear certain drawings of a cube as sometimes convex
and sometimes concave. What first, in the quantum mechanical approach,
had seemed to be fundamentally new and surprising, abruptly appears on
the contrary as endowed with a certain sort of universality, so of normality.
It leaps to one’s mind that:
* Any explicit account of a process of description, in so far that it is self-
contained, always includes a full specification of the action by which
the object-entity is introduced, as well as a full specification of the
action by which a qualification is obtained for this object-entity.
* Often these two actions are mutually independent.
* The introduction of the object-entity is sometimes achieved by creation
of this entity, while the operation of qualification, if it is a physical
process, always—in principle at least—changes the object-entity, and
sometimes radically, in which cases the relativizing consequences of one
or the other or both these epistemic actions, upon the development of
the process of description, have to be explicitly taken into account and
thoroughly analyzed.
128 M. MUGUR-SCHÄCHTER

For instance, think of a detective who is searching for material indica-


tions concerning a crime. What does he do? He usually focuses his attention
on a convenient place from the physical reality, say the theatre of a crime,
and there he first operates extraction of some samples (he cuts out frag-
ments of cloth, he detaches a clot of coagulated blood, etc.); or he might
even entirely create a test-situation involving the suspects, and insure reg-
istration by hidden apparatuses, of their behaviours. Only afterward does
he examine the gathered samples or the behaviours registered during the
test-situation. One can equally think of a biopsy for a medical diagnosis,
or an extraction of samples of rock operated by a robot on the surface
of another planet, and the subsequent examinations. In all these cases the
observer-conceptor—more or less radically—generates an object-entity that
did not pre-exist in the desired state or quantity, in order to qualify it later
by operations that are quite independent of the operation which generated
these entities. And in certain cases the operation of examination so rad-
ically changes the object-entity, that, if several different examinations of
this object-entity are necessary, also several replicas of it must be produced.
Furthermore, the obtained qualifications arise indelibly marked by a double
relativity: relativity to the way of generating the object-entity (this way can
simply exclude certain subsequent examinations), and also a relativity to
the sort of examination that was achieved.
The preceding considerations prompt the following two correlated
remarks:
In the first place, the nature and realm assigned by classical think-
ing, to the genesis of communicable knowledge, are misleading and shrinked.
The whole zone where mind actively constructs, out of pure factuality, the
very first forms of new communicable knowledge, is so deep-set that it re-
mained hidden beneath the two basic building blocks of all the current oc-
cidental languages, namely subjects and predicates. These do both suggest
available, pre-existing states of fact. Furthermore, the primordial creative
zone of conceptualization remained cut off also from most classical scien-
tific representations. Notwithstanding the well-known analises of Husserl,
Poincaré, Einstein, Piaget, and many others, not only classical logic and
probabilities, but also the set theory (hence most domains of modern math-
ematics), modern linguistic and semiotic, etc., take their start from a level
organized above language, by use of—quasi-exclusively—language: physical
operations are not considered. And factuality is widely supposed to sponta-
neously imprint—via language—upon passively receptive minds, informa-
tion concerning already existing, actual properties of pre-existing objects.
The active role is assigned quasi exclusively to the exterior factuality, not
to the mind. This attitude, in fact, is stronger and more general concern-
QUANTUM MECHANICS VS. RELATIVIZED CONCEPTUALIZATION 129

ing object-entities (typical grammatical subjects) than concerning qualifica-


tions (predications). Anyhow, globally, an attempt at an integrated and sys-
tematic representation of the emergence of individual object-entities and of
qualifications of these, by deliberate epistemic actions, and the way in which
these products get integrated into communicable concepts-and-language, is
still lacking. From “the other part of the mirror” where the biological struc-
ture of the man’s body is placed, the cognitive sciences are trying to ini-
tiate a representation of the sensorial bio-physiological processes involved
in phenomenal appearances and in conceptualization, by including into the
domain of investigation the inner volume delimited by a man’s skin. But if
this inner volume is excluded, then it is quantum mechanics which—for the
first time—suggests the possibility of, and the method for a most deep-set
attempt at a purely psycho-operational representation of the processes of
conceptualization: an attempt founded on the very first interplay of what is
called mind, with unknown factuality, and involving explicitly the descrip-
tional aims, the physical operations and devices, and the evolving stratum
of pre-existing conceptualization.
In the second place, the descriptional scheme explicated from the
epistemic strategy involved in quantum mechanics, is paradigmatic. It has
captured in it a certain sort of epistemic universality. Quantum mechanics
involves a particular embodiment of an extreme epistemic situation, namely
that which is realized when a communicable conceptualization is researched
concerning non pre-existing physical entities of which—a priori—only the
possibility is conceived, and which, if then effectively generated, emerge
non-perceivable. In such extreme circumstances one has been compelled to
a radically active, constructive attitude, associated with a maximal decom-
position of the global process. All the stages of the desired description have
had to be built out of pure physical factuality, independently of one another,
each one in full depth and extension: the severity of the constraints revealed
the most complete and explicit descriptional scheme where any other more
particular description must find lodging. In this sense the quantum mechan-
ical descriptional scheme possesses a universal epistemological value.
As soon as this universal value has been understood, one finds oneself
in possession of a starting point for specifying a convenient canonical form
of any description. Indeed such a canonical form must be precisely a com-
plete abstract structure with a maximally carved out capacity. It must be a
void form, a mould, able to offer an available, specific, and sufficiently large
location, for any possible stage of any possible descriptional process. In this
or that given description, one or more locations offered by this canonical
form might remain partially or totally non utilized. But this, if it happens,
will be known since the form will exhibit a labelled void of estimated am-
130 M. MUGUR-SCHÄCHTER

pleness. For instance, if I say “I consider what I see just in front of my


eyes and this is a red surface”, with reference to the maximally complete
descriptional mould drawn from quantum mechanics it will appear that in
this case the two canonically distinct descriptional actions, of generation of
the object-entity, and of qualification of this entity, have coalesced in the
unique act of “looking just in front of my eyes”, which both delimits and
qualifies the object-entity. So the location reserved for the stage of inde-
pendent generation of an object-entity remains entirely void in this case. It
will also be possible to estimate the magnitude of only partial voids and
to draw consequences. For instance, imagine the assertion “I plucked this
flower, I examined its morphology with a microscope, and the result is this”.
Comparison with the canonical mould brings forth that this amounts to a
description where the object-entity—as such—is introduced by an only par-
tially creative action—plucking a flower—while the act of examination might
only very little change the object-entity initially introduced in this way. So
in this case the two distinct locations reserved in the canonical mould in
view of a possibly radical creativity in both the stage of production of an
object-entity and in that of qualification of it, are both made use of, but
each one to only a very reduced degree. It follows that a classical treatment
(assuming the pre-existence of the object-entity as well as its invariance with
respect to the process of qualification) can be posited to produce a very good
approximation to the result that would be obtained by a complete canonical
treatment.

4. NORMS FOR DESCRIBING: THE METHOD OF


RELATIVIZED CONCEPTUALIZATION (MRC)

4.1. Preliminaries

Since 1982 I never ceased developing the method of relativized conceptual-


ization ([14,17,18])—let us denote it as MRC—founded on the generalization
of the descriptional scheme which I explicated from the quantum mechan-
ical descriptions. This method can be regarded as an attempt at a certain
“normation” of the processes of description of any sort, or in other terms, a
normation of the processes of communicable conceptualization.
Because of the descriptional relativitisations that are explicitly built
into it at each descriptional step, MRC withstands by construction the in-
sertion of false absolutes, thus warding off false problems or paradoxes. And
because it roots its constructions in physical factuality, at the lowest descrip-
tional level that can be reached, MRC furthermore withstands any gliding
into relativism:
QUANTUM MECHANICS VS. RELATIVIZED CONCEPTUALIZATION 131

MRC stands in polar opposition to what is called relativism.

It means confined, delimited, but strict precision of each descrip-


tional step, associated with free though guided choices of the way of connect-
ing the descriptional steps accordingly to the evolution of the descriptional
aim. Which insures controlled rigor throughout a progressive construction of
freely decided trajectories and nets of conceptualization, always indefinitely
open.
The main difficulty has been to find a way of escaping the impris-
onment inside the forms which current language, surreptitiously, imposes
upon thought. In all the preceding publications concerning MRC, in order
to achieve this liberation I made use from the start on of certain ideographic
symbolizations, but I never tried to achieve a mathematical formalization.
The ideographic symbolizations, however, have been felt by many to stay
in the way of a natural and full access to meaning. Therefore in this work
I adopt a different strategy. In a first stage I expose the nucleus of MRC in
usual language, trying to get through the stubborn implicit forms of thought
induced by the current usage of words, with the help of exclusively the re-
sources of the associations of words themselves (and of abbreviating literal
notations of words). In a second stage I give a summary of the ideographic
symbolization utilized in all the previous expositions of MRC, because it
permits a more suggestive and economic expression of certain basic con-
cepts and assertions. Finally, in a third stage I sketch out a mathematical
formalisation of the nucleus of MRC in terms of the theory of categories.7
This section is devoted exclusively to the nucleus of MRC. The way
in which the nucleus works will be illustrated in the subsequent Sec. 5, by
showing how it generates a deep and fully relativized unification between
the logical conceptualization and the probabilistic one.

7
The possibility of also another sort of mathematical formalization, more fit for calcu-
lations permitting numerical estimations—namely in terms of Hilbert-Dirac “individual”
vectors (i.e., not belonging to a vector-space)—will be found in the exposition of meta-
[quantum mechanics] (note 1). While in Sec. 5 it will become clear that the probably
most natural vocation of MRC is to yield a non-mathematical formal system comparable
to Russel and Whitehead’s Principia Matematica, but concerning conceptualization in
general instead of only logic.
132 M. MUGUR-SCHÄCHTER

4.2. The First Stage: a Presentation of MRC in Usual


Language

In what follows I formulate definitions (D), a postulate (P), principles (P),


conventions (C), and assertions which are called propositions (π) because
they are justified by “natural deductions” (indicated by the word “proof”
written between quotation marks in order to distinguish from deductions
inside a formal system). Each step is labelled by the symbol of its nature—
D, P, P, C, or π—followed by the ordinal of the step. There are 19 steps,
namely 15 definitions, 1 postulate and 3 principles. When a step is splitted
in sub-steps a sub-ordinal is added for each sub-step. A step is often followed
by comments.
I proceed by enumeration of the steps and sub-steps. The sequence
is interrupted by several intermediary titles which break the progression in
small groups each one of which concentrates upon a given purpose.

4.2.1. Preparation of the concept of relative description

D1. Consciousness functioning. The activity of an observer-


conceptor’s mind—called here consciousness functioning and noted CF—is
conceived to play a central generative role, acting on the exterior universe
and on the interior universe where it belongs, and there, in particular, also
on itself. This activity is regarded as the quintessence of the epistemic ac-
tor, irrepressibly anterior and exterior to any specified epistemic action. It is
an (the?) invariant among all the epistemic actions the observer-conceptor
is aware of, it is the tissue of his continuity, and each one of its products
becomes exterior to it as soon as it has been produced. It marks a mobile,
permanent and non removable cut—a ultimate cut—between itself and the
rest.
Comment. The Cartesian cut between res cogitans and res extensa
is second with respect to this mobile cut.

Throughout what follows CF is explicitly incorporated in the rep-


resentation. Thereby, from the start on, this approach breaks openly and
radically with the classical concept of objectivity. It introduces basically, in
a declared and systematic way, the supplementary representational volume
that is necessary for a non-amputated expression of the new concept of objec-
tivity in the sense of inter-subjective consensus, such as this concept emerged
from modern physics, from quantum mechanics and Einsteinian relativity.
That is, inter-subjective consensus founded on systematically extracted frag-
ments of pure factuality (quantum mechanics) and qualified by qualificators
explicitly constructed in order to express definite classes of relative obser-
QUANTUM MECHANICS VS. RELATIVIZED CONCEPTUALIZATION 133

vational invariance (Einsteinian relativity). Indeed both these constraints,


that are the core of modern physics, involve CF in a quite essential way.

D2. Reality. What is called reality is posited here to designate the


evolving pool—always considered such as it is available at the considered
time—out of which any given consciousness functioning either radically cre-
ates, or delimits, or only selects, object-entities of any kind whatever, phys-
ical or psychical or of a mixed kind. This pool will be indicated by the letter
R.
Comment. This non-restricted definition of “reality” refuses the
disputes on “existence” (do unicorns exist? does the number 3 exist? does
a class exist? etc.). It will appear that inside the present approach the in-
distinctions entailed by this absence of restrictions generate no difficulties;
that, on the contrary, they permit a posteriori clear definitions which so far
could not be reached in the approaches in which more specification has been
introduced at the start.

P3. The realist postulate. Throughout what follows is explicitly


postulated the existence—independently of any mind and of any act of
observation—of also a physical reality.
Comment. In the formulation of P3, as also in D1 and D2, the
specific designatum of the expression “physical reality” (that implies that a
sub-realm of what is called reality is considered), is assigned the status of a
primary datum. This however is only a starting point. In what follows the
general reflexive character of MRC will manifest itself, in particular, by the
fact that, progressively, a more constructed distinction between “physical”
reality and reality in general will constitute itself inside MRC.8

8
This specification takes into account concurrent remarks by Jean-Louis Le Moigne,
Michel Bitbol, Jean-Blaise Grize, and Gérard Cohen-Solal who—independently of one
another—argued that the concept of “physical reality” seemed to them neither clear nor
necessary in a context of the nature of MRC; that inside such a context this concept
should emerge. Furthermore, on H. Barreau’s opinion, speaking of “physical” reality might
erroneously suggest some confusing necessary connection with Physics, which the word
“empirical” would avoid. It will however appear that the crucial definition D14.3.1 of
a basic transferred description, as well as the preparatory points 8 to 13, are endowed
with significance exclusively with respect to what is usually called physical reality, while
with respect to reality in the general sense of D2—which includes, for instance,empirical
economic or cultural data, empirical aspects or components of what is called art, etc.—
the formulations from the points 8 to 14 are meaningless. So I simply do not know how
to avoid the assertion ab initio of P3 such as it is expressed above: such is the force of
language. On the other hand, throughout the points 8 to 14 the concept of physical reality
keeps acquiring constructed specificity. In this sense, a progressive specification of P3 does
emerge from the evolving MRC-context, as desired by the above-mentioned colleagues,
but it emerges on the basis, also, of P3 itself. So my final option is to conserve [D2+P3].
134 M. MUGUR-SCHÄCHTER

The posit P3 of existence of a physical reality might seem to be en-


tailed by D2, so redundant, but in fact it is not. Indeed, though everybody
agrees that what is called physical reality does contribute to the pool out
of which the consciousness functioning extracts object-entities to be stud-
ied, nevertheless the various disputes concerning “existence” of this or that
sort of object-entity (does Jupiter exist?) continue steadily. The associa-
tion [D2+P3] is intended as (a) a memento of the fact stressed most by
Descartes and recognized by the majority of the philosophers, that, in the
order of the emergence of knowledge, the assertion of the existence of phys-
ical reality cannot be considered to be primary with respect to the assertion
of the existence of subjective psychical universes (as classical physics might
seem to suggest) : the word “also” in the formulation of P3 is intended to
provocatively remind of this; (b) an explicit refusal of solipsism, on the other
hand; (c) an inclusion in what is called reality, of the concepts and systems
of concepts, of the behaviours, beliefs, social and economical facts, etc. (the
third world of Popper).

D4. Generator of object-entity and object-entity. The epis-


temic operation by which a consciousness functioning introduces an object-
entity will be regarded as an action upon R achieved by CF by the use of
a generator of object-entity denoted G. The spot (or zone, or the sort of
domain) from R where a given generator G acts upon R, is considered to be
an essential element from the definition of that generator, and which has to
be explicitly specified; it will be denoted RG . The object-entity introduced
by a given generator G will be denoted œG . For methodological reasons, a
one-to-one relation is posited between a given definition of a generator G
and the corresponding object-entity œG : that which emerges as the product
of a given G-operation, whatever it be, is called “the object-entity produced
by G and is labelled œG .

Comment. Any description involves an object-entity. Usually it is


considered that it suffices to name or to label this object-entity thus just
directing the attention upon it before it is more thoroughly examined. This
“linguistic” attitude is restrictive since not any conceivable object-entity
pre-exists available for examination. Therefore throughout what follows it is
required that the primordial epistemic action accomplished upon R which
brings into play the considered object-entity—as such—no matter whether
this action is trivial or not, be always indicated explicitly and fully.

For the moment it is sufficient to understand the qualification “physical” as pointing


toward anything involving an in principle definible amount of mass-energy. Then certain
non-physical entities, like “art”, etc., can involve physical aspects, while others, like the
concept of the number 3, do not.
QUANTUM MECHANICS VS. RELATIVIZED CONCEPTUALIZATION 135

A generator G of object-entity can consist of any psycho-physical way


of producing out of R an object for future examinations. Such a way involves
systematically some psychical-conceptual component, but which can com-
bine with concrete operations. A generator G can just select a pre-existing
object or on the contrary it can radically create a new object. If I point my
finger toward a stone I select a physical entity by a psycho-physical selective
gesture that acts in a non creative way on a physical zone from R (here
RG is the volume where the stone is located). If I extract from a dictio-
nary the definition of a chair I select by a non creative psycho-physical act,
an abstract conceptual entity materialized by symbols in a physical zone
from R consisting of the dictionary (so here R G ≡ dictionary). If I con-
struct a program for a Turing machine in order to examine the sequences
produced by this program, I bring into play a creative, instructional con-
ceptual generator of object-entity that acts on a zone from R containing
subjective and inter-subjective knowledge as well as material supports of
these. If, in order to study a given state of an electron, I generate it by using
some macroscopic device that acts on a place from the physical space of
which I suppose that it contains what I call electrons, I delimit a physical
object-entity, by a psycho-physical creative action. If now I apply the same
operation upon a mathematical theory, or upon a place from the physical
space where the vibrations of a symphony can be heard but the presence of
electrons is improbable, according to the definition D4 I am making use of
another generator, since it involves another zone RG , and, in consequence of
the one-one relation posited between G and œG , I delimit another object-
entity (interesting, or not, probably not, in this case). When I define by
words a new concept, as I am doing now, in order to later specify its be-
haviour, I produce a conceptual object-entity, by working, with the help
of a psycho-conceptual-physical creative generator, upon the spot from R
consisting of the reader’s mind.
The inclusion, in the definition of G, of the “zone” RG from R where
G is supposed to act, requires two important specifications:
(a) RG is not a qualification of the produced object-entity œG , ob-
tained by examining this object-entity in order to learn about it. It is a
condition imposed upon the operation of generation G in order to insure the
location of all the products of G, inside a pre-decided conceptual volume in-
dicated by some verbal label, “microstate”, “chair”, “program”, etc. In the
particular case of a selective generation like for instance pointing toward a
stone, this pre-posited conceptual volume where G has to act, might degen-
erate inside our mind into an identification with the physical location of the
object-entity œG , which has to be avoided). The methodological necessity
of such a pre-decided conceptual location will be fully understood later, in
136 M. MUGUR-SCHÄCHTER

the comment of the definition D14.3.1.


(b) The “zone” RG from R where G is supposed to act permits of un-
controllable fluctuations concerning what is labelled œG . The physical region
from R where I act in order to generate a given microstate of an electron, can
contain non perceptible and uncontrollably variable fields, etc.; the reader
of these lines can happen to be a 16 years old boy, or a mature intellectual.
These fluctuations entail an unavoidable non-predictability concerning the
effect labelled œG of an operation of generation of an object-entity. However
one should clearly realize that it simply is inconceivable to “entirely” im-
mobilize a priori the effect of G denoted œG : this would require to specify
“completely” RG . But such a requirement is both impossible (circular) and
unnecessary. One simply cannot start a process of representation of the way
in which descriptions, i.e., qualifications of any object-entities, emerge out
of R, by specifying, so qualifying R itself everywhere and for any time, and
also from any point of view. Such a circle cannot be realized. While the a
priori non-determination concerning the effect of the individual operations
of generation of an object-entity, is by no means an insuperable problem or
a difficulty. It simply is an unavoidable constraint that MRC is obliged to
recognize, include and control. The recognition of this constraint plays an
essential and very original role in the dynamics of conceptualization from
MRC. It brings into evidence one of the roots of human conceptualization
and it comes out to be intimately tied with a reflexive character of MRC, of
maximal a priori freedom, followed by a posteriori controls and restrictions.
It opens up the way toward a constructive incorporation (via the sequence
D14 of definitions of relative descriptions) of the fundamental fact called
“non-determination of reference” established by the deep analyzes of Quine
[10] and Putnam (cf. [11]), which marks the breaking line between factuality
and mere language.
Consider now the one-one relation posited between a given defini-
tion of an operation G of object-entity generation and what is labelled œG .
This relation is intimately tied with the above mentioned a priori non-
determination involved by RG , so also with the non-determination of ref-
erence. It is important to realize that no other relation could be uphold ab
initio. Indeed in general the object-entity labelled œG emerges still non qual-
ified from the standpoint of the subsequently intended examinations, if not,
in general its generation would be unnecessary for this aim. It can even
emerge still entirely inaccessible to direct knowledge of any sort, if G is a
radically creative and physical operation of generation (as in the case of
the microstate generated by most quantum mechanical operations of state-
generation). In these conditions what we called a one-one relation between a
given definition of an operation G of object-entity generation, and the mere
QUANTUM MECHANICS VS. RELATIVIZED CONCEPTUALIZATION 137

label œG , obviously cannot mean that the still unqualified replicas of œG are
all “identical” in some inconceivable absolute sense.
The one-one relation posited between G and œG amounts to just a
methodological pre-organization of the language-and-concepts, unavoidable in
order to be able to form and express a beginning of the desired representation
of a human conceptualization. Such a methodological pre- organization is,
by its nature, a formalizing step, like an algebraic rule.
Indeed if from the start on we imagined that G might produce some-
times this and sometimes something else, how would we speak of what it
produces, or think of it? We would have to re-label in only one way the
product entailed by a given definition of G, whatever it be, and thus we
would come back to precisely our initial choice of language and notation.
On the other hand, if we asserted a priori a “real” one-one relation between
G and what is labelled œG , we would thereby assert the sort of view that is
sometimes called metaphysical realism (a God’s Eye view, as Putnam puts
it), which would directly contradict the very philosophical essence of the
present approach. In the sequel, each time that some definite consequence
of this a priori choice of language will appear, we shall deal with it for that
definite case.
The explicitly methodological character of this constructive strategy
adopted in the definition D4, is a quite crucial step. It saves premature, void,
illusory questions and paradoxes that simply cannot be solved a priori. In-
stead, as it will appear, it brings forth a posteriori a clear, fully relativized
operational concept of “identity” that emerges progressively in π12, π13
and D14.1 and then is specifically defined in π18.1; which suppresses inside
MRC one of the most noxious false absolutes induced by current language.
And the relativization of the qualification of identity permits then immedi-
ately to show by π18.2 and π18.3 that MRC, inside its soma progressively
structured from the precedingly posited definitions, postulate and principles,
eventually entails a well-defined sort of minimality of the realist postulate
P3, initially posited without any further qualification. By this minimality
the “metaphysical realism” will appear to by organically rejected by MRC.

D5. Qualificators

D5.1. Aspect-view. Consider a grid for examination which, via


certain operations of examination performed on an object-entity œG , can be
a priori imagined to produce qualifications of this entity. Such a grid will be
called an aspect-view and will be denoted Vg . By definition Vg is structured
as follows:

- The qualifications that can be generated by Vg are contained inside a


138 M. MUGUR-SCHÄCHTER

semantic dimension called aspect and labelled globally by some index g


(which can take on any graphic form: another letter, a group of letters,
some other sign).
- The qualifications that can be generated by Vg are called g-
qualifications. The set of all the possible g-qualifications is allowed
to be arbitrarily rich but it is required to be finite, so discrete. Each
g-qualification is called a value k of the aspect g, in short a gk-value,
where gk—in one block—functions as only one index. The aspect g is
conceived to contain the corresponding finite set of gk-values, not to
identify with it.
- A gk value is permitted to be of either a physical or an abstract nature,
but it is required to be directly perceptible by the involved observer-
conceptor, via his biological senses and his mind.
- The aspect g is considered to be defined if and only if the specification
of its values gk is associated with also the explicit specification of an
effectively realizable modality—physical, or conceptual (in particular
formal), or mixed—for:

* Accomplishing the examinations—physical, or psychical or


conceptual—from the semantical dimension consisting of the as-
pect g.
* Expressing the results of these examinations in terms of “values
gk of the aspect g”, which amounts to the explicit specification
of certain coding-rules.

Any object, device or algorithm involved by the modality required


above, is to be included in the definition of the aspect g.
Comment. So, in contradistinction to the grammatical or logical
predicates, an aspect-view Vg is endowed by definition with a structure, and
with coding-rules which fix a finite “gk-language” consisting of operations,
signs, names, referents, and the stipulation of the relations between these.
This structure exhibits explicitly all the restrictions to which is sub-
jected an effectively realizable operation of qualification, that can be made
use of without incurring ambiguities. If these restrictions are not all satisfied
we simply are not in presence of an aspect-view in the sense of D.5.1.
Let us note that an order between the values gk of an aspect G is not
required but is permitted.
The distinction between an aspect g and the set of all the gk values
contained inside that aspect, takes into account the remarkable psycholog-
ical fact that any set of gk-values, even only one such value, as soon as it
is “conceptualized” (i.e., as soon as it ceases to be a mere “primeity” in
QUANTUM MECHANICS VS. RELATIVIZED CONCEPTUALIZATION 139

the sense of Peirce), generates in the consciousness a whole semantic dimen-


sion g (a genus) that exceeds this set and constitutes a ground on which
to place its abstract feet: every gk-value determines a location (a specific
difference) on this semantic domain g that grows spontaneously beneath it
(for instance, if gk labels the interior event toward which the word “red”
points, this event, when conceptualized, generates the carrying semantic di-
mension toward which the word “colour” points). We are in presence of a
fundamental law of human conceptualization that moulds logic, language,
and even metaphysics (the concept of “substance” is the semantic ground
on which are located the ways of existing of material systems, etc.). The
adopted definition reflects this law, on which it tries to draw the attention
of the cognitivistic approaches (what are the corresponding bio-functional
substrata?).
Finally let us also note that, by definition, an aspect-view Vg acts
like a qualifying filter: it cannot yield qualifications different from any cor-
responding gk-value.

D5.2.View. A grid for examination that consists of a finite but ar-


bitrarily large set of aspect-views, is called a view and is denoted V .
Comment. The complexity and the degree of organization of a
given view V are determined by the number of aspect-views Vg from V and
by the structures of the various sets of gk-values introduced by the various
involved aspect-views from V (number of gk-values, “position” (central, ex-
treme) of each set of aspect-values on the corresponding semantic dimension
g, existence or not of an order among the gk-values of a fixed aspect g, a
reference-gk-value (a gk-zero), etc.). In particular a view can reduce to only
one aspect-view or even, at the limit, to one aspect-view containing only
one gk-value on its semantic dimension g. There is nothing absolute in the
distinction between an aspect-view and a view: an aspect-view can be trans-
formed in a view by analysis of its aspect in two or more sub-aspects, and
vice-versa the set of distinct aspects from a view can be synthesized into a
unique aspect. This stresses that a view, like also a generator of object-entity,
is just a construct freely achieved by the acting consciousness-functioning
CF, in order to attain a definite epistemic aim.

D5.3. Physical aspect-view and view. Consider an aspect-view


Vg where the aspect g is physical and requires physical operations of ex-
amination of which the results consist of some observable physical effects.
Such an aspect-view will be called a physical aspect-view. A view containing
only physical aspect-views will be called a physical view (concerning this
language, see note 8).
Comment. This definition can be best understood per a contrario.
140 M. MUGUR-SCHÄCHTER

A mathematical or a logical view is not a physical view, though the involved


examinations do involve certain physical actions (writing, drawing, etc.),
because what is called the results of the examinations (not their material
expression) consists of concepts, not just of physical entities (marks on a
measuring device, for instance). (And of course, a physical view does not in
the least necessarily involve physics).

D5.4. Spacetime aspect-views. One can in particular form a space-


time aspect-view VET . Accordingly to Einsteinian relativity the double index
ET can be considered as one aspect-index g = ET where E reminds of the
current Euclidian representations and T stands for time. However the partial
aspect-indexes E and T can also be considered separately from one another,
setting g = E or g = T . The space-aspect E is associated with space-values
or “positions” that can be denoted E R ~ (setting a position vector R~ in the
role of the index k introduced in D5.1) and the time-values can be denoted
T t (setting a time parameter t in the role of k). Indeed though in general the
numerical estimations indicated by R ~ and t are not mutually independent,
nothing interdicts to symbolize separately the spatial position-value and the
time-value.
Infinitely many spacetime views can be constructed (by varying, in
the representations, the choice of the origins of space and time, of the units
for measuring intervals, the form and direction of the involved reference-
axes). Any spacetime aspect-view introduces an ordered grating of spacetime
values. This is a specificity with highly important epistemic consequences
([17] and Sec. 5.2 of this chapter), because it endows the spacetime views
with the power to strictly singularize an object-entity.

D6. Epistemic referential and observer-conceptor. A pairing


(G, V ) consisting of a generator G of object-entity and a view V , is called
an epistemic referential.
A consciousness functioning CF that endows itself with a given
epistemic referential is called an observer-conceptor and can be denoted
[CF, (G, V )].
Comment. A pairing (G, V ) is permitted to be entirely arbitrary a
priori. This is a methodological reaction to an unavoidable constraint: the
capacity of a pairing (G, V ) to generate meaning, can be examined only after
having considered that pairing. This particular methodological reaction is a
new manifestation of an already mentioned general reflexive strategy prac-
tised in MRC, of a tentative a priori approach that is entirely non restricted,
but is systematically followed by a posteriori corrective restrictions.
An observer-conceptor [CF, (G, V )] is the minimal epistemic whole
able to achieve epistemic actions in the sense of MRC: by itself an epistemic
QUANTUM MECHANICS VS. RELATIVIZED CONCEPTUALIZATION 141

referential (G,V) is not yet a closed concept, nor does it designate an active
entity. This concept becomes closed and activated only when it is associated
with the consciousness functioning CF that generated and adopted it.

D7. Relative existence and inexistence. Consider an a priori


pairing (G, Vg ). If an examination by the aspect-view Vg of the object entity
œG generated by G, never reveals to the involved observer-conceptor some
value gk of the aspect G, we say that the object-entity œG does not exist
(is not pertinent) with respect to the aspect-view Vg (or equivalently, that
Vg does not exist with respect to œG , or that œG and Vg do not mutually
exist)).9
Suppose now, on the contrary, an act of examination by the aspect-
view Vg of the object entity œG generated by G, that does reveal to the
involved observer-conceptor one or more values gk. In this case we say that
the object-entity œG exists with respect to the aspect-view Vg (or that Vg
exists with respect to œG , or that Vg and œG do mutually exist).
Comment. The definitions of relative inexistence or existence can
be transposed in an obvious way to one single value gk of an aspect g, or to
a whole view V .
The concepts of mutual inexistence or existence concern, respec-
tively, the general impossibility or possibility of the emergence of meaning,
as well as the intimate connection between meaning and descriptional aims,
which are induced by a tentative pairing (G, Vg ) or (G, V ). These concepts
are essentially semantical. They express the general fact—previous to any
qualification—that a given object-entity can be qualified only via the views
to the genesis of which it can contribute by yielding matter for abstraction.
Furthermore, the concepts of relative inexistence and existence permit to
cancel a posteriori, among all the initially only tentative pairings (G, Vg ) or
(G, V ) that an observer-conceptor has introduced, those which appear to be
non-significant; while the other pairings can be kept and put to systematic
descriptional work. The possibility of such a selection illustrates again the
general reflexive strategy of MRC: maximal a priori freedom followed by a
posteriori controls and restrictions.
The concepts of relative inexistence and existence have quite funda-
mental consequences, but with respect to which the classical conceptualiza-
tions are more or less blind. This generates various sorts of false problems
and paradoxes.

9
If one examined with the help of a voltmeter, a symphony by Beethoven, the operation
might never produce an estimation of a difference of electrical potential (accidents being
neglected). Of course during a more realistic sort of tentative research a mutual non-
pertinence can be much less apparent a priori than in this caricatured example.
142 M. MUGUR-SCHÄCHTER

P8. The Frame-Principle. I posit the following principle, called


frame-principle and denoted by FP.
Consider a physical object-entity œG that can be (or is conceived to
have been) generated by some definite physical generator of object-entity,
G. The frame-principle FP asserts the following:

- This entity œG does exist in the sense of D7 with respect to at least


one physical aspect-view Vg (D5.3) (if not the assertion of a physical
nature of œG would be devoid of foundation (content)).
- If the physical object-entity œG does exist in the sense of D7 with re-
spect to the physical aspect-view Vg , then ipso facto œG exists in the
sense of D7 with respect to also at least one view V formed by asso-
ciating Vg with a convenient spacetime view VET (it cannot exist with
respect to any such association, if only because the values gk of a given
aspect g can appear or disappear with respect to a given spacetime
view when the spacetime units are changed). But the object-entity
œG is non-existent in the sense of D7 with respect to any spacetime
view that acts isolated from any other physical aspect-view Vg where
g 6= ET : The spacetime views are frame-views which, alone, are blind,
they cannot “see” anything.
- According to what precedes what is called “physical spacetime” can-
not be regarded as a physical object-entity œG . Indeed the assertion
posited in the first part of this principle does not apply to what is
called “physical spacetime”: The designatum of this expression itself,
considered strictly alone, is non-existent in the sense of D7 with re-
spect to any physical aspect-view Vg where g 6= ET , and it is equally
non-existent with respect to any association of such a physical aspect-
view, with a spacetime aspect-view. In this sense: If spacetime were
regarded as a physical object-entity we would need spacetime where
to locate it and thus we would be drawn into indefinite regression.

What is called “physical spacetime” is—itself—only the locus of all


the possible spacetime frame-views (referentials), the genus of these.
It is the conceptual volume where physical entities, facts or aspects,
can be assigned spacetime specifications which, if this is desired, can
be numerically defined by the use of spacetime referentials.

Comment. The frame principle FP adopts, transposes in terms


of MRC, and specifies, the Kantian conception according to which man is
unable to conceive of physical entities outside physical spacetime, that he
introduces as a priori “forms of the intuition inside which he casts all his rep-
resentations of physical entities. FP isolates and stresses certain particular
QUANTUM MECHANICS VS. RELATIVIZED CONCEPTUALIZATION 143

implications of this Kantian conception which so far seem to have remained


insufficiently noticed by physicists. Namely that any mature and normal
human being, by the nature of his consciousness functioning, as soon as he
perceives or even only imagines a phenomenal appearance which he connects
with what he conceives to be a physical entity œG , ipso facto introduces more
or less explicitly:

(a) A spacetime frame-aspect-view VET (the observer-conceptor’s body


tends to yield—vaguely—the intuitive origin, the units, and—
variable—directions of the axes, whereas in the technical or scientific
approaches these are explicitly and freely specified in a precise and
stable way, in mathematical, integral or differential terms).
(b) At least one aspect-view Vg , where g is a physical aspect different
from VET , relative to which the considered physical entity œG does
exist in the sense of D7, and the values gk of which he combines with
the value-indexes E R ~ and T t of the spacetime aspect-view VET (in
mathematical terms, with the spacetime coordinates yielded by VET ).
J. Petitot ( [4], p. 216) writes concerning Kant’s conception of space
and matter:

“As quality (not as quantity any more), matter is filling of


space. This filling is very different from a mere ‘occupation’
(anti-Cartesianism). It is a dynamical and energetical process
characteristic of the substantial ‘interiority’ of matter.”

In P8 the necessity of the presence of at least one physical aspect g


different from the space or time aspects is a way of expressing the
presence of the matter that fills spacetime and of asserting that any
phenomenal manifestation to human minds, if it does not stem from
the inner universe, stems from this matter, not from spacetime itself.
(c) With the help of a spacetime frame-view alone, in the strict absence of
any other sort of physical aspect-view Vg (colour, texture, whatever),
man is unable to perceive or even imagine a physical entity. He simply
is unable to extract it from the background of exclusively spacetime
frame-qualifications which, by themselves, act only as elements of a
grid of reference inserted in an abstract, void container labelled by
the words “physical spacetime”. By themselves these elements of a
grid of reference act exclusively as potential land-marks that can be
“activated” only by the values of some other aspect g 6= ET .

The assertion that the designatum of the words “physical spacetime”


cannot be treated itself as a physical (object-)entity—probably obvious for
144 M. MUGUR-SCHÄCHTER

most physicists—is introduced here explicitly mainly in order to emphat-


ically block certain very confusing ways of thinking induced in the minds
of non-physicists by the verbal expressions by which the physicists use to
accompany their relativistic formalizations: these verbal expressions suggest
that what is currently called spacetime would itself possess this or that
metric; while in fact any spacetime metric is just assigned by construction
to this or that spacetime frame-aspect-view, on the integral level or on the
infinitesimal differential level, on the basis of some definite (even if implicit)
descriptional aim (this is discussed in the last chapter of this work).
The frame principle is endowed with a strong formalizing power which
imprints its marks upon all the scientific representations of physical reality,
inside physics as well as inside “abstract” mathematics.

C9. Conventions. In order to take explicitly into account the frame


principle FP we introduce the following conventions:

- Any view V considered in order to examine a physical object-entity


will contain a spacetime aspect view VET and one or more physical
aspect-views Vg .
- The aspects denoted g are always different from the spacetime aspect
ET.

P10.The principle of individual spacetime mutual exclusion.


Consider a physical object-entity œG corresponding to a physical generator
G. Let V be a physical view with respect to which œG does exist in the
sense of D7, involving two distinct physical aspect-views Vg1 and Vg2 as well
as a spacetime view VET (accordingly to C.9). The principle of individual
spacetime mutual exclusion posits the following:

- Any physical examination involved by V quite systematically changes


the state of the examined physical object-entity œG , even if only to
a degree which in this or that context can be neglected: the state
of a physical object-entity is not a stable datum with respect to an
act of physical examination (in informatics one would say that it is a
“consumable” datum).
- If, when performed separately on different replicas of œG , the examina-
tions involved by Vg1 and Vg2 can be shown to cover different spacetime
domains—the referential and the origins for spacetime qualifications
being kept the same—which involves that they change differently the
state of œG —then it is not possible to perform both these two sorts
of examinations simultaneously upon a unique replica of œG produced
by only one realization of G (the word “individual” from the denom-
QUANTUM MECHANICS VS. RELATIVIZED CONCEPTUALIZATION 145

ination of P10 refers to this crucial unicity of the involved replica of


œG ).

If the type of impossibility specified above manifests itself, the two physical
aspect-views Vg1 and Vg2 6= Vg1 are said to be mutually incompatible. In the
alternative case Vg2 and Vg1 are said to be mutually compatible.
Comment. It is probably possible to draw back P10 to other still
more basic spacetime mutual exclusions (an attempt has been made in [17]B,
p. 290). But here, for simplicity, we start from the formulation P10 because
it is more immediately related with the consequences pointed out in the
sequel.
The quantum mechanical principle of “complementarity” can be re-
garded as the realization of P10 for the particular category of physical object-
entities consisting of states of microsystems. This brings into clear evidence
the often only obscurely perceived fact that complementarity in the sense of
quantum mechanics has an—exclusively—individual significance: indeed two
mutually incompatible quantum mechanical measurements can be simulta-
neously realized on two distinct replicas of a given microstate (object-entity),
and if this is done two distinct and useful pieces of information are obtained
in a quite compatible way [13]. But this brings already up on a statistical
level, and there what is called the mutual incompatibility of two physical
aspect-views is not manifest any more. What is impossible indeed is only the
simultaneous realization of two mutually incompatible quantum mechanical
measurements upon one given replica of the considered microstate.

The concept of incompatibility of two physical aspect-views is defined


only with respect to one individual replica of some given object-entity:
it is not intrinsic to these physical aspect-views.

This is of crucial importance from a logical point of view; cf. Sec. 5.1.2.

Π11.Proposition. Consider a physical object-entity œG correspond-


ing to a generator G and a physical view V with respect to which œG does
exist in the sense of D7. In general, in order to perform upon œG all the
operations of examination corresponding to all the different aspect-views Vg
from V , it is necessary to realize a whole set of successions [(one operation
of G-generation of œG ), (one operation of Vg -examination of that replica of
œG )] (in short [G.VG ]) containing (at least) one such pair for each physical
aspect-view Vg from V .
“Proof”. In order to achieve examinations of œG via mutually in-
compatible physical aspect-views Vg from V , the operation G of generation
of œG has to be repeated (the time parameter being re-set to its initial value
146 M. MUGUR-SCHÄCHTER

t0 , as in sport-measurements, in the repetitions of chemical or physical exper-


iments, etc.) and paired successively with these incompatible aspect-views.
Comment. This, though an obvious consequence of P10, is highly
non-trivial by itself. It is important to know explicitly that the achievement
of complex examinations of an object-entity involving “consumable” char-
acters, entails in general the condition of reproducibility of all the involved
pairs [G.VG ] (either in succession or in simultaneity), thus involving a whole
set of replicas of the involved sort of object-entity œG . (The proposition π11
and its “proof” admit of generalization to also certain conceptual referentials
(G, V )).

Π12. Proposition. Consider a physical object-entity œG correspond-


ing to a given generator G, and one given physical aspect-view Vg with re-
spect to which œG exists in the sense of D7. When a succession [G.Vg ] is
repeated a big number N of times (the time parameter being re-set for each
pair to its initial value tœ ) or when it is simultaneously realized on a big
number of replicas of the object-entity œG , it is not impossible that the same
observable gk-spacetime-values be found in each instance; in such a case one
can say that an individual qualificational N -stability has been obtained. But
in general this does not happen; in general the N obtained gk-spacetime val-
ues are not all identical, notwithstanding that in each realization of a pair
[G.VG ] the operations G and Vg obey strictly the same defining conditions.
“Proof”. This follows per a contrario: to posit a priori that the
results produced by repeated realizations of a given succession [G.Vg ] are all
identical “because” in each pair both G and Vg obey the same specifications,
neither follows with necessity from the previously introduced definitions and
principles, nor could it be found a posteriori to be always factually true. To
show this last point it is sufficient to produce a counter-example. Consider
an object-entity generator G which acts by definition on a zone RG from R
consisting of a piece of land, and that delimits there the object-entity œG
consisting of a definite area of one square kilometer. Let Vg be an aspect-view
(structured accordingly to D5.1 and C9) that permits to establish the aspect
G '[association of mean-colour-value-and-space-position over a surface (any
one) of only one square meter]: inside the epistemic referential (G, Vg ), two
distinct realizations of the succession [G.Vg ] in general yield two different
results, even though both G and Vg satisfy each time to the same operational
commands.
Comment. Notice that if an individual qualificational N -stability is
found for a given succession [G.VG ], this by no means excludes the possibility
that in another series of N 0 repetitions (with N 0 bigger than, or equal to, or
smaller than N ) no individual stability will be found any more.
QUANTUM MECHANICS VS. RELATIVIZED CONCEPTUALIZATION 147

Furthermore, and this is more important, if for a given object-entity


œG corresponding to a given generator G, an individual N -stability with
respect to the examinations by a given aspect-view Vg is found, this does
by no means involve that for the same object-entity œG but another aspect-
view Vg0 with g 0 6= g one will find again some individual stability for some
big number.

The individual stability of the qualifications of an object-entity œG


or the statistical character of these, are relative to the qualifying
aspect-view Vg .

It is of the utmost importance to realize that—quite generally—a


generator G of a physical object-entity being fixed by some operational def-
inition of it, it would even be inconceivable that for any association of G
with some aspect-view Vg , the results of repetitions of the corresponding
sequence [G.VG ] shall all be identical: that would be a miracle in so far
that absolute identity—independent of the considered aspect-view Vg , i.e.,
for any tried aspect-view Vg —has never been observed concerning a physi-
cal object-entity which—factually—is always endowed with strict singularity
(this probably holds even for a conceptual object-entity, like, say, the num-
ber 5, if its mental correspondent in a given mind is considered). As for
“identity” in absence of any view—which, as many do in fact surreptitiously
and vaguely imagine, would mean identity of œG with itself from one re-
alization of G to another one, not of the qualification of œG via Vg when
the succession [G.Vg ] is repeated—it is but an illusory concept tied with the
quest for an impossible absolute objectivity of the thing-in-itself. (The psy-
chological difficulty encountered for realizing this stems from the physical,
“exterior” nature supposed for œG , which surreptitiously inclines to posit
that—like œG itself—the qualifications of œG also exist independently of
any observer-conceptor, as “intrinsic properties” of œG ).
The above considerations bring back to the only methodological
meaning which can be a priori assigned to the one-one relation posited
between G and œG , and, correlatively, they bring back to also the roots of
the non-determination of reference.
(Notice how all the preceding assertions acquired inside MRC a “nat-
ural” deductive character (i.e., outside any formal system) manifesting the
formalizing essence of the features with which we progressively endow this
approach. Which is a quite non-trivial feature of MRC).

Π1 3.Proposition. Given an epistemic referential (G, Vg ) where both


G and Vg involve physical operations, in general no stability at all is insured
for the gk-spacetime values obtained by repeated or multiple realizations of
148 M. MUGUR-SCHÄCHTER

the succession [G.Vg ], neither on the individual level of observation, nor on


the statistical one.
“Proof”. If only a maximal, an individual N -stability is considered,
i.e., identity of all the N groups of observable gk-spacetime values corre-
sponding to N realizations of a succession [G.Vg ], then π13 becomes a mere
repetition of π12, hence the “proof” of π12 still works. But suppose that
no individual N -stability has been found, i.e., that a whole statistical dis-
tribution of dispersed triads of gk-spacetime-values has been found. Then it
still remains a priori possible that a big number N 0 of repetitions of a series
of a big number N of repetitions of the succession [G.Vg ](N 0 6= N in gen-
eral) shall bring forth, when N 0 is increased toward infinity, a convergence
in the sense of the theorem of big numbers, of the relative frequencies of
occurrence, in the mentioned statistical distribution, of the dispersed triads
of gk-spacetime-values. In this case one can speak of a probabilistic (N, N 0 )-
stability. However, up to some given arbitrary pair (N, N 0 ) of big numbers,
it might appear by experiment that in fact this second possibility does not
realize either, even though G and Vg have been found to mutually exist in
the sense of D7. Nothing excludes such a situation, neither some previous
MRC-assumptions, nor the empirical experience. If this negative situation
does realize indeed, then only two solutions are left: either one continues the
search with pairs of increasingly bigger numbers N ,N 0 , or one stops at some
given pair (N, N 0 ) and announces a posteriori that, even though G and Vg
do mutually exist in the sense of D7, their pairing (G, Vg ) has nevertheless to
be (N, N 0 )-cancelled from the subsequent conceptualization, because, while
no individual N -stability has been observed, this pairing does not generate a
probabilistic (N, N 0 )-stability either; tertium non datur because apart from
an individual or a probabilistic stability, no other sort of still weaker stability
has been defined so far (in Sec. 5.2 this question is treated more thoroughly).
In short, for any given pair of big numbers (N, N 0 ), it is quite possible that
no stability at all be found for the results of repeated successions [G.VG ].
Which establishes π13.
Comment. The “proof” of π13 does by no means exclude the pos-
sibility that, if the succession [G.VG ] does produce a probabilistic (N, N 0 )-
stability, another succession [G.Vg0 ] with G the same but with Vg0 6= Vg , shall
produce qualifications that are endowed with some individual N -stability, or
with no stability at all, neither probabilistic nor individual:

The existence of a probabilistic stability of the qualifications of a given


object-entity œG is relative to the qualifying aspect-view Vg just like
the existence of an individual stability. The nature—individual or
probabilistic—of the stable qualifications of a given object-entity œG ,
is relative to the qualifying aspect-view Vg just like the existence of
QUANTUM MECHANICS VS. RELATIVIZED CONCEPTUALIZATION 149

stable qualifications.

4.2.2. The normed concept of relative description

D14. Relative description

D14.1. Relative description of a physical object-entity. Con-


sider an epistemic referential (G, V ) where: G is a physical generator that
generates a corresponding physical object-entity œG ; V is a physical view
with m aspect-views Vg with respect to each one of which œG does exist
in the sense of D7; and, as required by P8 and C9, V contains also a
spacetime view VET introducing an ordered spacetime grating (D5.4). Fur-
thermore consider, for each Vg from V , a big number N of realizations of
the corresponding sequence [G.VG ], in simultaneity or in succession, the time
parameter being re-set at the same initial value t0 for each realization of a
sequence [G.Vg ].
Suppose first that, when the succession [G.Vg ] is realized N times, for
each aspect-view Vg from V , identical outcomes of the corresponding config-
uration of gk-spacetime-values are obtained, i.e., only one same “individual”
result appears N times. We shall then say that an N -individual outcome has
been obtained (the reference to this N is necessary because nothing excludes
that for another sequence of successions [G.Vg ] some dispersion be found).
The set of N -individual configurations of gk-E R-T ~ t-values corresponding
to all the m distinct aspect-views Vg from V , constitutes in the abstract
representation space of V ordered by the spacetime grating introduced by
~ t- values. This “form” will be called an N -
VET , a definite “form” of gk-E R-T
individual relative description, with respect to V, of the physical object-entity
œG , (in short an individual relative description) and it will be indicated by
the notation DN /G,œG ,V / to be read “the description relative to the triad
G, œG ,V and to N ” (in current usage the index N , supposed to be big, will
be dropped). The individual relative description D/G,œG ,V / defined above
can also be regarded as the set of all the individual one-aspect-relative-
descriptions D/G,œG ,Vg / with Vg ∈ V .
Suppose now that, when the various successions [G.Vg ] with Vg ∈
V are realized N times, not all the successions [G.Vg ] are found to reproduce
~ t-values; that at least for one
identically one same configuration of gk-E R-T
Vg ∈ V (not necessarily for all) the corresponding succession [G.Vg ] pro-
duces a whole set Sgi = {cgi } of mutually distinct, dispersed configurations
~ t-values, (with i ∈ I and I a finite index-set, to preserve
cgi of gk-E R-T
the finitistic character of this approach); but that, for any succession [G.Vg ]
which produces dispersed results, when N is increased toward infinity, the
relative frequency n(cgi )/N of occurrence of each configuration cgi ∈ Sgi
150 M. MUGUR-SCHÄCHTER

converges toward a corresponding probability pgi . In these conditions each


configuration cgi ∈ Sgi will be called an elementary-event-description cor-
responding to the succession [G.Vg ] with Vg ∈ V and it will be denoted
Dp(gi) /G,œG ,Vg /. The epistemic referential (G, V ) will be said to produce a
probabilistic relative description of the physical object-entity œG which will
be denoted Dp /G,œG ,V /.10
Comment. The definition D14.1 is the core of MRC. It finally as-
signs a significance to what has been called a physical object-entity œG . A
significance which, though it is relative to a view V and in certain “basic”
conditions that will be specified in D14.3.1 is far from being fully “satisfac-
tory”, nevertheless is now quite definite and endowed with communicability.
Whereas G alone cannot systematically insure for “œG ” a significance dis-
tinct from just the conventional label “effect of a realization of G”, because
the results of G might emerge still entirely non perceptible.

D14.1.1. Reference and relative meaning. In any case of qual-


ificational stability, individual or probabilistic, we shall say that œG is the
reference (or referent or designatum) of D/G,œG ,V / while D/G,œG ,V / is
the meaning of œG relatively to V .
Comment. It thus appears that the initial methodological assertion
of a one-one relation between a given definition of an operation G and its
result labelled œG , does not hinder the subsequent construction of all the
necessary specifications. On the contrary, it founds them.

The following is noteworthy:

The condition of existence of individual or probabilistic stability of the


outcomes of the successions [G.Vg ], with respect to repetitions of these,
pre-supposes the possibility to achieve arbitrarily many successions
[G.Vg ], for all the Vg ∈ V.

This is a strong restriction. But when it is insured it extracts out


of temporality the concept of “description” founded upon it and it puts
it directly on highways of communicability where reference, meaning, and
objectivity in the sense of intersubjective consensus, can most immediately
be attained. Furthermore, it sets a standard with respect to which relaxing
generalizations can be now defined.
10
This definition of a probabilistic description is incomplete and simplifying. It will be
thoroughly reconstructed and completed in Sec. 5.2. A more ancient but full treatment
can be found in Ref. [18]. At this stage of the development of MRC we are obliged to
introduce it in this unachieved form, as a provisional support for essential distinctions
that cannot be postponed.
QUANTUM MECHANICS VS. RELATIVIZED CONCEPTUALIZATION 151

D14.2. Two generalizations of D14.1

D14.2.1. Relative description of a non-physical public object-


entity. Let us suppress in the definition D14.1 the restriction to physical
generators, while excluding generators that act on only one individual inner
universe (there, in general at least, the sequences [G.Vg ] cannot be repeated
(in succession or in simultaneity) and so the condition of stability of their
results cannot be insured). Thus relaxed, the definition D14.1 enlarges to
object-entities from the non physical but public, exterior reality (economical,
social) for which the repeatability of sequences [G.Vg ] and the condition of
stability of their results still may happen to make sense. The new sort of
description obtained in this way will be called a relative description of a
non physical and public object-entity and it will be indicated by the notation
(N P P ).D/G,œG ,V /, in short (N P P ).D.
Comment. The generalization D14.2.1 holds in particular concern-
ing any already accomplished description in the sense of D14.1, selected itself
as a new, always conceptual object-entity, to be examined in a subsequent
description via some new view. Thereby:

The definition D14.2.1 opens up specifically and explicitly the whole


crucially important sub-realm of R consisting of constituents of a sta-
bilized communicable conceptual reality where, in particular, “logic”
is located.

In the case of non-physical object-entities that admit of a description


in the sense of D14.2.1, any reference to the frame-aspect of (”physical”)
space can obviously be dropped, and so the obtained relative description
amounts to a “form” of only gk-time values. If moreover it appears that the
considered description can be regarded to be independent also of time values,
(as for instance in the study of a fixed formal system), the reference to the
frame-aspect of time can be equally dropped. (For instance, the dependence
on time cannot be dropped for the relative description pointed toward by
the verbal expression “this theory is true”: The truth-value yielded by the
examination of the object-entity consisting of a theory, via the aspect-view
Vg where g = truth, does depend on the structure of knowledge (informa-
tions, understanding, modalities of verification, etc.) available to the acting
observer-conceptor at the considered time; on the contrary, for the relative
description indicated by the verbal expression “the sum of the angles of a
Euclidean triangle is 180◦ ”, the time dependence can be dropped). Con-
sider then a relative description where both the space qualifications and the
time-qualifications can be dropped. If no one among the involved aspects g
introduces by its own definition an order (cf. D5.1), this description consists
152 M. MUGUR-SCHÄCHTER

of one or several non-ordered but stable configurations of gk-values. What


does this mean? It means that the involved non-ordered configurations are
characterized by some correlations, which are stable with respect to repeti-
tions of the sequences [G.Vg ] permitted by the view V , i.e., a given gk-value
is found to be associated with this or that other g 0 k 0 -value (g 0 6= g or k 0 6= k
or both), always, or never (which is as strong a correlation as always), or
with this or that probability.
D14.2.2. Relative testimony. Take again as a starting point the
strong definition D14.1, and suppress now in it both the restriction to only a
physical generator of object-entity and the condition of repeatability of the
sequences [G.Vg ] for the Vg from V . What becomes of D14.1? It reduces to
a mere set of “qualifications” generated by a definite epistemic referential.
Indeed as soon as an epistemic referential (G, V ) is given and the condition
D7 of mutual existence is satisfied for the pair (G, V ), qualifications via V
can arise for the object-entity œG produced by the generator G. From now
on any structure of such qualifications will be called a relative testimony and
will be denoted θ/G, œG , V /, in short θ.
Comment. The generalization D14.2.2 of D14.1 gives a definite
status inside the MRC-language to all the qualifications of unique object-
entities of any nature. In the case of physical object-entities, uniqueness is
often intimately connected with spacetime singularity, in particular with the
principle P10 of individualizing spacetime mutual exclusion. This will come
out to have a surprising importance in the identification of the characteristics
of the deepest stratum of an MRC-logic (Sec. 5.1.2).
Furthermore, D14.2.2 introduces in the MRC-language all the quali-
fications of psychical events from the inner universe of a conceptor-observer.
This is a huge inclusion that lays down a foundation for the future research
of a clear connection in MRC-terms, between introspective reports and neu-
rological facts. Which might lead to comparability of the MRC requirements
on this sort of connection inside the framework of important new views on
body versus mind, like those of Edelman [7], Changeux [6], Damasio [8], and
more generally, in connection with the whole avalanche of results continu-
ally produced in the cognitive sciences. Thereby the problems of reference
and truth that haunt this vast recent domain might find the conceptual
framework for a guided approach.
Finally, the relative testimonies in the sense of D14.2.2 permit to
take into consideration the historical descriptions, the poetical ones, etc.
For these the fundamental concepts of reference and truth still remain wide
open for discussion and for methodological organization.

D14.3. Basic transferred relative descriptions. In what follows


QUANTUM MECHANICS VS. RELATIVIZED CONCEPTUALIZATION 153

we finally shall touch and transpose in quite explicit and generalized terms,
the fundamental epistemological innovation specifically implied by quantum
mechanics.

D14.3.1. Basic transferred relative descriptions of a physical


object-entity. Consider a relative description in the sense of D14.1 where:

- The generator consists of a repeatable physical operation and it pro-


duces a physical object-entity that cannot be perceived directly by
man. Such a generator will be called a basic generator and will be
denoted G(o) .
- The object-entity produced by a basic generator G(o) will be called
a basic object-entity and will be denoted œ(o) (a simplified notation
standing for (œG(o) )(o) ).
- The view able to draw phenomenal manifestations out of a basic
object-entity is necessarily such that the phenomenal content of each
gk-value of each involved aspect g consists of features of a material de-
vice for gk-registrations, biological or not, but which always is different
from the studied object-entity, these features emerging as “marks” pro-
duced by the interactions between the registering-device and replicas
of the considered basic object-entity. These marks acquire significance
by their coding in terms of values gk of the aspects from the acting
view. A view of the just specified kind will be called a basic transfer-
view (in short a basic view) and will be denoted V (o) . The aspect-views
(o)
from V (o) will be called basic aspect-views and will be denoted by Vg .
- The epistemic referential (G(o) , V (o) ) will be called a basic epistemic
referential.
- A relative description in the sense of D14.1, individual or probabilistic,
achieved with a basic generator and one basic transfer-aspect-view
(o)
Vg , will be called a basic transferred relative aspect-description and
(o)
it will be denoted D(o) /G(o) , œ(o) ,Vg /.
- A relative description in the sense of D14.1, individual or probabilistic,
achieved with a basic generator G(o) and a basic transfer-view V (o)
(o)
involving at least two mutually incompatible basic aspect-views Vg1
(o)
and Vg2 , will be called a basic transferred relative description (also, in
short, a basic description or a transferred description) and it will be
denoted D(o) /G(o) , œ(o) , V (o) / (in short D(o) ).
- A basic transferred description D(o) /G(o) , œ(o) , V (o) / is posited to
characterize observationally the involved basic object-entity œ(o) ,
which means by definition that it is posited that no other operation
154 M. MUGUR-SCHÄCHTER

of generation (G(o) )0 6= G(o) can be found which, associated with the


same basic view V (o) , produces the same basic transferred description.

Comment. It is difficult to fully grasp the meaning and the impor-


tance of the concept of basic transferred relative description. But it is crucial
to grasp it fully. Indeed it is by this concept that MRC penetrates beneath
natural language and the forms of thought involved by it, establishing a def-
inite and rather complex relation between conceptualization and physical
factuality. Therefore I shall comment on it in detail, even redundantly.
To begin with, let us stress that a basic physical object-entity pro-
duced by a basic physical operation G(o) , if furthermore this sort of object-
entity has never before been qualified via any transfer-view V (o) whatever,
emerges still entirely unknown in terms of the knowledge researched con-
cerning it specifically, notwithstanding that the operation of generation G(o)
does singularize it out of the whole of physical reality. Indeed—factually –
the result labelled œ(o) is entirely “specified” by G(o) , it is “defined”, since
it is made available for any possible subsequent examination and, accord-
ing to D14.3.1 and to the posited one-one relation between the operation
G(o) and its result œ(o) , it can be deliberately reproduced. More. Factually,
each such result emerges from the operation G(o) that produced it, fully in-
dividualized, it lies on a level of zero-abstraction, still filled with its whole
untouched concrete singularity. Which no language whatever could never
do because we generalize as soon as we speak: full singularity is unspeak-
able. But—consequently in fact—this result produced by G(o) alone, not yet
followed by an operation of examination, is individualized in another man-
ner than that in which knowledge concerning it—specifically—is researched;
namely in an only factual, physical sense, not an already conceptualized,
qualifying sense. It is true that the specification of the generation operation
G(o) involves necessarily some position of a pre-decided conceptual space
of qualification (tied with the “zone” RG from R where G(o) is supposed
to act (D4 and comment on it). By its definition G(o) drops its products
inside this pre-decided conceptual volume. That what is labelled œ(o) is
pre-constrained to emerge inside this or that spacetime domain where G(o)
acts, it is produced so as to correspond to some definite verbal designation
(”a manifestation of stellar life”, or “a state of a microsystem”, etc.). In
this sense G(o) and its result labelled œ(o) might be considered to never be
“purely” factual. But:

The preliminarily posited conceptual volume where the operation G(o)


drops its products, cannot be equated to the new knowledge that is
researched concerning these products. The elaboration of this new
researched knowledge is the task left by construction for examinations
QUANTUM MECHANICS VS. RELATIVIZED CONCEPTUALIZATION 155

to be achieved subsequently upon the already produced œ(o) , by this or


(o)
that basic aspect-view Vg that exists in the sense of D7 with respect
to, non-specifically, anything lying inside the pre-decided conceptual
volume where G(o) drops all its products.

It is important to realize that the specification of the operation G(o)


of generation of an object-entity must contain a conceptual receptacle at-
tached to the physical action involved by G(o) ; a conceptual receptacle to be
lowered with this action into the depths of pure as yet non-conceptualized
physical factuality, in order to receive inside it the results of the operation
G(o) so as to be able to hoist them up into the stratum of the concepts-and-
language. This is an unavoidable condition because only a receptacle made
of concepts-and-language can hoist up into the thinkable and speakable a
lump of pure factuality. A macroscopic operation G(o) can be itself shown,
taught, repeated, and also said, pointed towards by words. But if nothing
thinkable and speakable were posited concerning what G(o) produces, which
by hypothesis is not perceivable, then this, the product, even if factually it
has been produced, would simply stay out of conceptualization. While hu-
man mind, in order to be able to think about a non perceivable thing, needs,
not only to have labelled it by a repeatable operation of generation and by a
notation, but furthermore to have endowed it with some initializing concep-
tual status, with at least some approximate preliminary speakable location
inside the unending and infinite-dimensional space of concepts.11
But of course a basic description D(o) does not indefinitely produce
an object-entity œ(o) that is still unknown, specifically and precisely in the
desired terms. Knowledge about œ(o) is a subjective and moving charac-
ter. Think of a basic operation of generation G(o) that is repeated by the
observer-conceptor X after it has founded for him the desired knowledge
(o)
concerning œ(o) , via some basic operations of examination Vg : Then even
though œ(o) is generated by the same generator G(o) and emerges beneath
the level of the directly observable by man, it is nevertheless already known
to a certain extent by X (while for another observer-conceptor it can be
strictly unknown, even if the knowledge acquired by X has been made
socially available in public registration devices (apparatuses, catalogues,
books, etc.). The only objective (inter-subjective) and perennial features
of a “basic” description D(o) and of what is here called a “basic” object-
entity œ(o) stem from the constant character of the involved referential, a
“basic” referential (G(o) ,V (o) ) where G(o) works on the physical factuality
11
It was Evelyne Andreewsky who, by repeated questions and remarks, motivated me
to specify how, exactly, the pre-existing conceptualization and the descriptional aims act
upon the extraction of new knowledge out of as yet unconceptualized physical factuality.
156 M. MUGUR-SCHÄCHTER

and V (o) is a transfer-view as specified in the definition D14.3.1: it resides in


the fact that what is called a basic description D(o) consists by definition of
exclusively features imprinted upon registering devices that are all different
from the studied object-entity œ(o) .
Consider now the following question which is fundamental for the
MRC treatment of reference: Does indeed the definition D14.3.1 of a basic
description open up a way toward a communicable characterization of—
specifically—the basic object-entity œ(o) ? The final posit from D14.3.1 con-
(o)
cerns this question. Consider a basic aspect-description D(o) /G(o) ,œ(o) ,Vg /
(o)
(the basic view consists of only one basic aspect Vg ). In this case it seems
clear that D(o) does not yield a characterization—individual or probabilistic,
no matter, but specifically and isolately—of what is labelled œ(o) , since it
points toward observable manifestations brought forth by interactions be-
tween œ(o) and a material device for gk-registrations. Which changes what
was labelled œ(o) (P10) and produces perceivable results that depend on the
device for gk-registrations as much as of œ(o) . But what about a “binocular”
basic description D(o) where the basic view V (o) consists of two mutually in-
(o) (o) (o)
compatible basic aspect-views Vg1 and Vg2 6= Vg1 ? In quantum mechanics,
for the particular case of a basic object-entity that is a state of a microsys-
tem, it is (implicitly) admitted that, together, two quantum mechanical
descriptions of a same microstate via two mutually incompatible quantum
mechanical views, characterize that microstate. Which means only that no
other operation (G(o) )0 6= G(o) of generation of a microstate can be assumed
to yield both these same two quantum mechanical descriptions. The final
posit from D14.3.1 generalizes inside MRC the above-mentioned quantum
mechanical implication. It would be satisfactory of course to base this posit
upon a general constructed argument (for instance a reductio ad absurdum).
But so far I have failed to find one. So I introduce the condition as just
a supplementary security for the solidity of MRC). This completes now on
the observational level the methodological posit from D4 according to which
a given operation of generation of an object-entity is assumed to always
produce the same object-entity. The necessity of a complement of this type
can be best understood per a contrario. In the absence of any phenomenal,
specific, normed, communicable set of qualifications associated specifically
with what has been labelled œ(o) , one would have to regard “œ(o) ” as just a
label that labels nothing distinct from this label itself. Then speaking and
thinking of “what has been labelled œ(o) ” would be only a void sophistic
trick, amounting to arbitrary implicit postulations.12 We would be obliged
to admit that pure factuality and human communicable knowledge stay for
12
Putnam’s thought experiments concerning the non-determination of reference [11]
are very suggestive.
QUANTUM MECHANICS VS. RELATIVIZED CONCEPTUALIZATION 157

ever apart from one another. But this just does not happen. Quite on the
contrary, our capacity to adapt to the environment and the technical pow-
ers that we are able to acquire manifest continually the astonishing, even
miraculous agreement between human knowledge and factual being, attest-
ing intimate couplings and transmissions which somehow manage to emerge
between them.
The posit from D14.3.1 incorporates into the MRC-representation the
assertion of a definite way in which a basic object-entity produced by a basic
generator G(o) , can be conceived to be captured inside pure physical factual-
ity and then hoisted up into the conceptual net of inter-subjective knowledge:
it is that what produces a pair of sets of mutually incompatible observable
manifestations which - accordingly to the final posit from D14.3.1—cannot
0
be obtained by the use of any other operation G(o) 6= G(o) .
At a first sight the concept of a basic transferred description might
seem very particular, and too radical. But in fact it possesses absolute pri-
ority and non restricted generality inside the order of cognitive elaborations.
Quite universally, any object-entity corresponding to any generator, if it did
reach the consciousness of an observer-conceptor, then it reached it first by
some transferred descriptions. We remain unaware of this because usually
the phenomenal appearance of the gk-values involved in these transferred de-
scriptions stems from marks imprinted directly upon the biological domains
of sensitivity of the observer’s body which act at the same time as gener-
ators of object-entity and as views in the sense of MRC. So the involved
epistemic referentials are of a nature which, with respect to the general
MRC-descriptional mould, is particular and degenerate (cf. the global com-
ments on D14 and the comments on D19.4, 5.1.1, and 5.1.2). This entails the
following effects which occur all at the same time and beyond any control
of logical consistency:
(a) It hides the transferred character of the marks.
(b) It inclines toward assigning systematically a passive role to the mind,
in its interactions with physical factuality. The mind is supposed to just
receive marks irrepressibly imprinted upon the sensitive apparatuses
of the body by incessant streams from the physical factuality. (How
far one is thus kept from realizing the possibility and the universal
methodological value of two radically distinct epistemic stages which,
in general, have to be both active during a deliberate achievement
of “unnatural” transferred descriptions, like those on which quantum
mechanics throws light!).
(c) It pushes surreptitiously toward ontological absolutizations. Indeed
one encounters severe difficulties to realize that the (various) trans-
ferred descriptions of this chair, which my consciousness functioning
158 M. MUGUR-SCHÄCHTER

achieved spontaneously by the help of my biological views (involv-


ing the eyes, the nervous system, the ears and fingers, etc.), cannot,
without contradiction, be identified with “the-way-in-which-the-chair-
in-itself-really-is”; that nothing, never, will be able to prove that this
or that model of a chair “exists” independently of any perception,
of any view. More, that such an instinctive hope contradicts both
philosophy and logic, since in the absence of any view the very con-
cept of description, and even that of merely an isolated qualification,
simply vanishes (cf. D14.2.2, D19.1, D19.2). It is really hard to with-
stand the irrepressible trend toward identification of our spontaneous
modelizations stemming from descriptions transferred on the human
biological registering devices, with ontological credos that float on self-
contradicting assemblages of words, alike to Magritt’s tree that floats
with its roots in the air. Kant, Poincaré, Einstein, Husserl, Quine, Put-
nam, have founded famous analyses on the explicit recognition of this
fact.

But, and this is noteworthy, as soon as the transfer-view from a considered


basic transferred description D(o) does not directly involve the biological hu-
man terminals—the nearest and which in fine cannot be eliminated—as soon
as the transfer-view V (o) from D(o) involves marks registered on devices that
are exterior to the observer’s body (as it happens indeed for microstates), it
suddenly becomes quite clear that a basic description D(o) itself constitutes
a constructed intermediary object-entity which relays the access of the basic
a-conceptual object-entity œ(o) , to the observer-conceptor’s consciousness-
functioning; that phenomena are not always independent of aimed volition,
that they are not always just psycho-physical facts which emerge sponta-
neously, but might have to be planned and produced by method. Then, like
in quantum mechanics, the two distinct and mutually independent stages
involved in a transferred description—the stage of generation of an object-
entity œ(o) , and the subsequent stage of creation of observable manifesta-
tions drawn from œ(o) by interaction with gk-registering devices—appear
as obvious. Their active and deliberate character strikes the mind, and the
invaluable normative value of the concept of basic transferred description
can be fully understood.
The basic object-entity œ(o) from a transferred description D(o) roots
this description directly into the physical factuality. Correlatively the trans-
ferred description D(o) achieves for the involved basic object-entity œ(o) a
very first passage from pure physical factuality, into the domain of communi-
cable knowledge. It yields for it a first communicable form, a first observable
expression that points communicably toward the involved object-entity. So
the basic transferred descriptions are the local zero-points of the chains of
QUANTUM MECHANICS VS. RELATIVIZED CONCEPTUALIZATION 159

conceptualization, in the following sense. Each basic transferred description


D(o) starts from a conceptual situation where, though some conceptual en-
vironment of the basic object-entity œ(o) (genus, etc.) always is more or less
explicitly posited a priori (at least via the definition D4 of G(o) ), nevertheless
nothing is known concerning œ(o) specifically.
The very first stratum of communicable knowledge available at any
given time consists of the basic transferred descriptions achieved up to that
time, not of just phenomenal appearances in the Kantian sense.

The transferred descriptions are the channels through which as yet


non semantized but semantizable factual matter, is adduced into the
domain of the inter-subjectively semantized. The “scientific legaliza-
tion of phenomenal appearances” in Kant’s sense (2.3) begins by the
construction of transferred descriptions, of which D(o) yields a form
that is normed. Which amounts to a formalization of the structure of
the connections between knowledge and Being.

This is a quite fundamental contribution of MRC to epistemology. It


separates the volume of the known in two essentially different strata. Indeed
the whole rest of the available knowledge consists only of subsequent devel-
opments of this first—evolving—stratum of transferred descriptions which
operate the very first connections between Being and knowledge. Namely
developments consisting of spacetime modelizations which endow the basic
transferred descriptions with the features required by the frame-postulate
P8, thus insuring for them an “intelligibility” of which initially they are
devoid; and then, a non limited succession of complexifications (or general-
izations, etc.) of these spacetime modelizations (cf. D.19 and all the involved
discussions).
I add a last remark concerning the concept of basic transferred de-
scription. From the viewpoint of MRC the quantum mechanical descriptions
of microstates appear now as just particular instances of transferred descrip-
tions of physical entities: the strategy of quantum mechanics, once identified
explicitly, brings into evidence an example of the universal way in which
the conceptualizations are rooted into pure physical factuality, and, for this
example, it displays all the stages of the rooting. MRC recognizes the uni-
versality of this rooting and extends it to any sort of physical factuality,
re-expressing it in general and normalized terms.

D14.3.2. Basic description of a psychical object-entity?


Notwithstanding important difficulties (the non-repeatibility of the succes-
sions [G(o) .V (o) ], it might turn out to be possible to forge a useful concept of
basic description of “psychical basic object-entities œ(o) ” (by some combi-
160 M. MUGUR-SCHÄCHTER

nation of testimonial descriptions θ in the sense of D14.2.2, with “biological


basic transferred descriptions”). Thereby I mean a conscious but not yet
conceptualized psychical object-entity, a primeity in the sense of Peirce that
emerges in the acting observer-conceptor’s interior universe, and, though
perceived, is still entirely unknown, non-qualified (A. Damasio ( [8]) has
elaborated a very subtle structure of concepts-and-facts concerning events
of this sort). Think for instance of all the feelings of mere existence of an inner
fact of which one becomes suddenly aware strictly without knowing as yet
explicitly what and how they are, so a fortiori without understanding them;
think of the genuine research conducted by Proust in order to identify the
subjective meaning of such feelings; think also of the psychoanalytic meth-
ods which deal with features as if transferred upon behavioural “devices”
(reactions, ways of acting, feelings) by interactions between a hypothetical
entirely unknown inner configuration, and various accidental or systemat-
ically arising exterior circumstances; this hypothetical inner configuration
is precisely what the therapies try to first somehow delimit “operationally”
(by analyses of dreams, associations, etc.)—even if by creating it—and then
to interpret, qualify, and control or suppress. The obtained description is
then in a certain sense precisely what seems to deserve being called a basic
relative description of a basic psychic object-entity.
It is however clear that for the moment these are just conjectures.
The central concept of basic transferred description has an indisputable
pertinence only with respect to physical object-entities.

Global comment on the definitions D14. Finally, let us now


consider globally the whole set of definitions D14 and make some comments
on the general concept of relative description.
The general notation D/G,œG ,V / stresses that any description that
is normed in the sense of MRC brings into play a triad G,œG ,V to which it
is essentially relative: this is the general descriptional mould induced from
quantum mechanics and required now for any description, whether it is
transferred or not. The first location from this triad is the place reserved for
an epistemic action, the generation of an object-entity, which up to now has
quasi systematically been ignored, because the canonical basic transferred
descriptions where the generation of an object-entity plays a separate and
active key role, were ignored. Indeed for a description that is not transferred,
the operation of generation of the desired object-entity is often accomplished
without any difficulty, in a natural or even implicit way (think of descrip-
tions of conceptual entities, for instance, the definition of the concept of
“table” that pre-exists implicitly inside my mind). While when the trans-
fers occur in—directly—the biological sensorial apparatuses (views, in the
QUANTUM MECHANICS VS. RELATIVIZED CONCEPTUALIZATION 161

sense of MRC), the involved view V acts also like a generator G which just
selects out of R an object-entity, namely the field of perceptibility of V ,
and—simultaneously—also qualifies this object-entity: we can symbolize by
G(V ) such a generator of a view and by (G(V ), V )) the corresponding epis-
temic referential. In this case the existence of a generator of object-entity is
still deeper hidden than in the preceding case. This highly degenerate and
so wide-spread natural situation contributed strongly to the lasting occul-
tation of the fundamental role of principle of the operations of object-entity
generation. Quantum mechanics, for the first time and only implicitly, made
a separate use of the operations of generation of object-entity, which per-
mitted to become aware of their general and fundamental epistemological
importance.

The generator of object-entity remained the big omission of the gram-


mars, the logic, and of all the approaches that involve the processes
of conceptualization.

This is why the question of reference has raised insuperable problems: the
basic object-entities are only surreptitiously drawn into the natural basic
descriptions (the degenerate ones being produced in a reflex way via the bio-
logical sensorial apparatuses), with the status of a present but non specified
reference. The problem of identifying a posteriori of what this reference con-
sists, starting from the already achieved description, has stubbornly resisted
solution.
But accordingly to MRC, an operation of generation of object-entity
is always involved, even if in a non separated and implicit or reflex way.
By construction, any relative description D/G,œG ,V / is, itself, dis-
tinct from the generator, the object-entity and the view involved by it, to all
of which it is conceptually posterior; it qualifies only the object-entity which
it concerns, not also the generator and the view of which it makes use, nor
itself, globally. As for the generator and the view, these are by definition
distinct from one another, often by their content, but in any case by the role
held during the process of description.

In the definition of a relative description the three notations G,œG ,V


designate three descriptional roles, three descriptional functions, not
the nature of the entities to which these roles are assigned in the case
of this or that particular relative description.

And all these three roles are systematically played in any relative
description, even if an actor cumulates distinct roles, or plays a role superfi-
cially, or both. For instance, if I say ““red” is a too poor expression, better
162 M. MUGUR-SCHÄCHTER

say “colour of blood””, the first proposition expresses verbally a relative de-
scription D/G,œG ,V / where “red”, though grammatically it is an attribute,
holds the role of the object-entity œG (generated by use of a generator G
which is a selector acting upon the spot RG from R indicated by the word
“colour”), while “poor” is placed in the role of the view V . But if I say “my
cheeks are red”, “red” plays the role of the view. So the structure required
by the definition D5.1 of an aspect-view, is only a necessary condition for
acting as a view, but this condition does not hinder a view in the sense of
D5.1 to act also in the role of an object-entity (like in the above first ex-
ample) or in the role of a generator G(V ) of object-entity that generates its
field of perceptibility by interaction with R.

According to MRC no operation or concept possesses intrinsically a


fixed descriptional role.

In each descriptional act, the descriptional roles are assigned by the


acting consciousness functioning, and in general these roles change from one
description to another one. When a natural description is examined in order
to compare it to the MRC norms, the first step is to examine what plays
the role of object-entity, what the role of generator, and what that of view.
A description D/G,œG ,V / is a piece of constructed normed meaning which,
essentially and explicitly, is relative to the epistemic actions that achieved
the semantization asserted by it. Any asserted meaning bears inside it the
genetic structure designated by the sign D/G,œG ,V /, but it can include this
structure in a more or less implicit, truncated, malformed way. Whereas
in the normed form D/G,œG ,V / all the three involved roles G,œG ,V are
explicitly indicated, each one at its own location and following the genetic
order of the corresponding epistemic actions. They are to be treated as void,
available, labeled rooms that have to be filled up in a check questionnaire
to which any achieved or envisaged description must be subjected.
The distinction, inside a relative description D/G,œG ,V /, between
the relativity to the operation G of object-entity generation of which the role
is to produce an object-entity, and the relativity to this object-entity œG
itself of which the role is to bear subsequent qualifying examinations, is one
of the most subtle and important features of MRC. In particular it preserves
from the very strong inertial tendency induced by classical thinking, to forget
that as soon as an entity is regarded as playing in a description the role of
object-entity, ipso facto a corresponding epistemic action of generation of
object-entity has produced it as such, implicitly or explicitly, even if this
entity somehow pre-existed and so has only had to be selected as object-
entity, not to be radically created as such. The importance of a normed
memento of this fact will fully appear in Secs. 5.1 and 5.2.
QUANTUM MECHANICS VS. RELATIVIZED CONCEPTUALIZATION 163

The association, in any relative description D/G,œG ,V /, between a


one-one relation G-œG and the requirement for D of, indifferently, either a
strong individual stability or an only probabilistic one, is intimately related
with the impossibility, for mere language as well as for mere notations, to
grasp and capture the factual individualities, neither in an absolute sense nor
in only a relativized sense (cf. π12, its “proof” and the comments). Umberto
Eco remarks: “The tragedy comes from this that man speaks always in a gen-
eral manner about things which always are singular. Language names, thus
covering the non transcendable evidence of individual existence“ [19]. Indeed
each predicate (view) is general, and no conjunction of a finite number of
predicates can ever exhaust the open infinity of the possible qualifications
of a physical object-entity.
The concept of relative description is selective. It does not admit in-
side the class delimited by it, illusory descriptions where one of the three
roles G, œG , V is not played at all. Consider for instance the famous illusory
description “this is a lie” (or “I am a lie”)” where the word “this” (or “I”)
masks the absence of specification of the operation G of generation of object-
entity, so also the absence of specification of the object-entity œG itself. In
fact, there simply is no object-entity at all. This blocks any further concep-
tual development. Indeed, previously to any research of a truth-qualification
of the description, one finds oneself in a situation of impossibility to decide
concerning the mutual existence in the sense of D7 between the involved
object-entity œG —nonspecified—and the involved view V . If this primary
non-decidability concerning the a priori possibility of meaning were some-
how (unimaginably) overcome, it would manifest itself later in the form,
also, of a paralysis of any attempt at a metaqualification of the relative
proposition founded on this illusory description via the values gk =true or
gk =false of a meta-aspect-view g = empirical truth (cf. DL.2 and DL.3 in
Sec. 5.1.2).
When descriptions that violate the MRC norms, are reconstructed
in a normalized way, the paradoxes stemming from them disappear. There
is no need for this to introduce levelled languages of logical types, the ill-
ness is cured locally by the normed reconstruction of only the considered
description.
But nothing hinders to generate (select) as an object-entity any nat-
ural description excluded by MRC, and to characterize its incapacities or
specificities by reference to the MRC-norms. In this sense the methodological
selectivity of the concept D/G,œG ,V / by no means constitutes an a priori
pauperisation of the ensemble of descriptions that can be studied inside
MRC.
Finally, the general concept of relative description, by its various
164 M. MUGUR-SCHÄCHTER

realizations, permits to discern definite categories inside the realm of the


problem of reference and of meaning, and a dégradé of proposed solutions:
the definitions D14.1, D14.2.1, and D14.3.1 introduce, for the corresponding
circumstances, what might stand as a solution or be completed to become
one; the definition D14.2.2 suggests a possible approach concerning some of
the circumstances to which it applies, while others are isolated as the most
problematic; finally, the non achieved definition D14.3.2 concentrates in it
definite questions and suggestions.
Like the one-one relation between a given generator of object-entity
and the corresponding object-entity, like the definition of relative existence
and then the frame principle P8, the concept of relative description with the
three roles involved by it, is an act of (qualitative) formalization involving
a methodological essence.

4.2.3. Cells of relative description. Chains of descriptional cells.


Non-reducible complexification of the conceptualization

P15.The Principle of separation. Since any one relative description


D/G,œG ,V /, whatever its complexity, involves by construction one generator
of object-entity, one object-entity, and one view, all well defined, as soon as
some change is introduced in the actor designated for holding one of the
roles from the triad G,œG ,V, another description is considered.

By a methodological principle called the principle of separation and


denoted PS, this other description must be treated separately.

Comment. Any human observer-conceptor, in presence of reality,


is condemned to parcelling examinations. The successivity inherent in hu-
man mind, the spatial confinements imposed by the bodily senses—whatever
prolongation is adjusted to them—and the absence of limitation of what is
called reality, compose together a configuration which imposes the fragmen-
tation of the epistemic quest. MRC reflects this situation in the relativity
of any one description, to one triad G,œG ,V . Indeed the relativity to one
triad G,œG ,V specifies, but also limits the capacity of one given relative
description to generate further information.

Relativization, limitation, and precision, are tied to one another in


an unseparable way. They constitute together an indivisible whole that
withstands relativism.

On the other hand, any fragment generated out of reality in order to


play the role of an object-entity, admits of an infinity of kinds of examina-
tions. Moreover any examination achieved on this object-entity, raises the
QUANTUM MECHANICS VS. RELATIVIZED CONCEPTUALIZATION 165

question of the appearance of its result via this or that view with respect to
which this result exists in the sense of D7, or the question of the relations
of this result to descriptions of other object-entities, etc., thus multiplying
the conceivable subsequent object-entities and examinations. These confine-
ments and these endless and changing vistas call forth haste and panics of
the mind which entangle in knots of “paradoxes” and block the understand-
ing. So they also block the further development of the started conceptual-
ization. The limitations imposed by each specified description are flooded
by the implicit fluxes of the rush toward more conceptualization. Without
being aware of this, mind yields to whirls of implicit interrogations which
generate a subliminal tendency to fluctuate between different operations of
generation of an object-entity and different views; a tendency to work out
simultaneously several different descriptions. But as soon as the elaboration
of several different relative descriptions is simultaneously tried, the various
involved generators of object-entity, object-entities and views, are offered a
ground for oscillation. And then the oscillations actually happen, because it
is very difficult to perceive them, so a fortiori to hinder them. So the dif-
ferent descriptions that are simultaneously entered upon, get mixed, and in
general none of them can be achieved. Their interaction coagulates nonsense
that stops the conceptualization.
The principle of separation hinders such coagulations. It requires the
conceptualization, by method, to be achieved by explicit separation in mu-
tually distinct, successive, closed, cellular descriptional steps.
In particular the principle of separation PS surveys the saturation
of a description. It rings the bell as soon as the descriptional capacities of
a started description must be considered to have been exhausted, because
all the qualifications via the view chosen for acting in that description, of
the object-entity corresponding to the generator chosen for acting in that
description, have been already realized by performing a big number of repeti-
tions of all the successions [G.Vg ]available in that description. PS announces
that once this has been done, the descriptional cell potentially delimited by
the chosen epistemic referential (G, V ) has been saturated with actualized
qualifications; that from now on any attempt at obtaining new information
inside this same epistemic referential, either is useless or it manifests the
surreptitious intrusion of another generator of object-entity, or of another
view, or both; that—to avoid stagnation, paradoxes or infinite regressions—
one has to stop this intrusion or mixture, by identifying the new epistemic
referential that weighs with subliminal pressure upon the consciousness func-
tioning, and by putting it explicitly to work in its own turn, separately.
The systematic application of the principle of separation plays, in
the development required by MRC for a process of conceptualization, a
166 M. MUGUR-SCHÄCHTER

role similar to that conveyed by the sign “.” or the word “stop” in the
transmission or writing down of a message; or else, a role similar to that
played in algebra by the closure of a previously opened parenthesis. It is a
formalizing requirement of the nature of a rule of calculus.
Any process of conceptualization that is normed accordingly to MRC,
is clearly divided in a sequence of localized descriptional cells, and thus it
develops by systematically renewed local frameworks, under systematically
renewed local control.
While the tests of mutual existence (D7) detect the a priori im-
possibilities to construct meaning, the principle of separation permits to
avoid any stagnation—illusory paradoxes, infinite regressions—throughout
the processes of development of meaning. The concepts of mutual inexistence
and the principle of separation cooperate to prevent sources of unintelligi-
bility, and also to detect and suppress them.
The principle of separation possesses a remarkable capacity of orga-
nization of the conceptualization. This assertion will find many illustrations
in the sequel of this work.

D16. Relative metadescription. The principle of separation re-


quires descriptional closures and new starts. These entail the necessity of an
explicitly and fully relativized concept of metadescription prescribing how
to transcend “legally” an already saturated description.
Consider a precedingly achieved relative description to which the or-
der 1 is assigned conventionally: D(1) /G(1) ,œ(1) , V (1) / (in short D(1) ; and
instead of œG we write œ, to simplify the graphism). Consider a genera-
tor that selects D(1) as a new object-entity œ(2) , denote it G(2) and call
it a metagenerator (or a generator of order 2 relative to D(1) . So we have
œ(2) ≡ D(1) . Consider also a view involving aspects of order 2 with respect to
which D(1) does exist in the sense of D7 (for instance the aspect of empirical
truth of D(1) , or else some aspect of relation inside D(1) /G(1) ,œ(1) ,V (1) /, be-
tween the various gk-spacetime qualifications produced by the examinations
of œ(1) by the initial view V (1) , etc.; call it a metaview (or a view of second
order) relative to D(1) and denote it V (2) . The description which is relative
to the triad G(2) ,œ(2) ,V (2) will be called a metadescription (or a description
of order 2) relatively to D(1) and it will be denoted D(2) /G(2) , œ(2) , V (2) /
(in short D(2) ).
The same denomination and notation are conserved if (a) G(2) selects
as a new object-entity œ(2) not only D(1) considered globally, but further-
more it includes in œ(2) also separate elements from D(1) /G(1) , œ(1) , V (1) /
specified explicitly (G(1) , or œ(1) , or V (1) , or two or all three of them) which
permits then to introduce in V (2) aspects of relation between such an el-
QUANTUM MECHANICS VS. RELATIVIZED CONCEPTUALIZATION 167

ement, and the global result D(1) to which it has contributed. Or if (b)
(1) (1) (1)
G(2) selects a whole set {D1 , D2 , ...Dm } of previously achieved relative
descriptions (with an explicit reconsideration, or not, of elements from these
descriptions), in which case D(2) is relative to all these descriptions. In this
way a very free and rich concept of normed relative metadescription is in-
troduced.13
Comment. The definition D.16 can also be applied to D(2)
thus leading to a metadescription D(3) of order 3 relatively to D(1) and
of order 2 relatively to D(2) , etc. In this way it is possible for any
consciousness-functioning CF to develop unlimited descriptional chains
D(1) , D(2) , ...D(j) ...D(n−1) , D(n) of hierarchically connected relative descrip-
tions of successive orders j = 1, 2, ....n—with an arbitrary origin denoted
D(1) —in each one of which the involved metaview can contain all the de-
sired pertinent new meta-aspects of order n.
So in general the order of a description is not an absolute, it labels
the place where this description emerges inside the considered chain of con-
ceptualization, while a chain can be started conventionally by these or those
previously achieved descriptions to which the order 1 is assigned.

But a basic transferred description can only have the minimal con-
ceivable order, no matter in which chain it is involved. Therefore this
non-conventional minimal order will be denoted by 0, to distinguish
it from any conventional initial order 1.

And any chain, if it has first been conventionally started with already
previously achieved descriptions to which the order 1 has been assigned, can
always be later completed downward until a basic transferred description
13
Here we can go back to the important distinction from the note 5 between “objec-
tual” qualifications—call them “objectities”—and “state”-qualifications. The objectities
are (relatively) stable qualifications that apply in an invariant way to a whole class of
evolving states, thereby definig the “object”, in the current sense, that assumes this or
that state. So according to this language the term object-entity labels only a descrip-
tional role in the sense of the general comment of D14, while “object” in the current
sense means “endowed with some objectities”: inside MRC these two words should not
be confounded. For instance, the state-qualifications called position, momentum, energy,
etc., can vary or evolve from one state to another one, thereby introducing an infinite
class of states of a definite sort of “object“ labelled, say, “electron”, which, inside a conve-
nient metadescription (with respect to this whole class of states), is characterized by the
metaqualifications consisting of the numerical values obtained (with some given system
of unities) for objectities like mass, charge, spin, that are the same inside the whole class
of what is called “states of electrons”. These objectities however can themselves change
by creation or annihilation of the corresponding object, and when the conditions for such
changes are realized they can be regarded as states of some more general object (at the
limit, of what is called field or energetic substance). In this way the language introduced
here can organize conveniently various hierarchies of degrees of abstraction.
168 M. MUGUR-SCHÄCHTER

is identified which roots the chain into pure factuality. Thereby the chain
hits an absolute end (or equivalently, it finds an absolute beginning), which
entails a corresponding re-notation upward, of all the successive orders of
the involved descriptional cells. But a given relative description can belong
to different chains that meet in it (it can be a node of the web of chains
of conceptualization); so, regarded as a cell from distinct chains, a same
description can have different orders: though all the basic descriptions are
absolute zeros of descriptional chains, a given non-basic description from a
chain possesses only relative orders with respect to its various zero-points.

Nevertheless, since the zero order of a transferred description is an


absolute, the feature of being a metadescription, or not, is an absolute
if transferred descriptions constitute the origin used as reference.

This amounts to the remark (rather obvious a posteriori ) that:

The (open) set of all the possible relativized descriptions falls apart
in just two (evolving) layers: (a) the layer of transferred descriptions
of physical basic object-entities which, by definition, are not them-
selves previously achieved descriptions, and (b) the layer of metade-
scriptions in the absolute sense, i.e., of descriptions of object-entities
consisting of previously achieved descriptions.14 Both layers have an
evolving content.

Through the first layer, the prime matter for the elaboration of mean-
ing is drawn into conceptualization, and inside the second layer the ba-
sic meaning produced in the first layer undergoes abstract transformations
which progressively elaborate indefinitely complexified meanings.
It is essential to note that in any chain, for each passage from
a descriptional level n to the following level n + 1, the new epistemic
referential to be used (G(n+1) , V (n+1) ) is freely decided by the acting
14
However it is curious to note that there are various sorts of rooting of a basic object-
entity, into pure factuality: the objectual manifestations of a basic object-entity, in the sense
of the note 13 can be conceived (not known, just imagined) to be tied with pre-existing
“own” features of this basic object-entity (cf. D19) which, though unknown, are always the
same. In this sense, a basic object-entity which is a priori researched as located inside the
genus labelled micro-object (i.e., is researched exclusively via objectual manifestations) is
thereby a priori endowed with a rooting into pure factuality which is less hidden than that
of a basic object-entity researched a priori as located inside the genus labelled microstate,
because it is posited to reach the level of observability by just a time-invariant coding
transposition, not by the coding of the effects of a (measurement) evolution produced by
the processes of examination. These remarks amount to the assertion of various possi-
ble deliberately chosen depths of the rooting of a transferred description, into physical
factuality.
QUANTUM MECHANICS VS. RELATIVIZED CONCEPTUALIZATION 169

consciousness-functioning CF, as an expression of his own (evolving) de-


scriptional curiosities-and-aims, such as these emerge at any given time from
his own biological, temperamental, and social-cultural background: it is the
consciousness-functioning CF who, step by step, chooses the “direction”
of the descriptional trajectory drawn by the succession of the cellular but
connected descriptional closures D(n−1) , D(n) , D(n+1) ,.... which, according to
[P15+D16], produce the indefinite progression of a hierarchical chain started
by conventionally initial conceptual descriptions D(1) or by absolutely initial
basic descriptions D(o) .
So—as long as no method or algorithm is found for determining auto-
matically, as a function of some definite parameters, a new epistemic referen-
tial, when a passage from a description to a metadescription (with respect to
it) has to take place—a descriptional chain remains a concept that cannot be
absorbed in the concept of computation. And even if such an algorithm were
specified, furthermore also the determination of the parameters on which the
new referential depends should emerge automatically: accordingly to what
criteria? Etc. The subjective successive descriptional aims play a decisive
role in the representation of the processes of conceptualization offered by
MRC. But on the other hand, the structure assigned by MRC to the con-
ceptualized, namely the structure of a web of chains of increasingly complex
relative descriptions, is a (qualitatively) formalized structure, involving def-
inite methodological rules and conventions.

This brings clearly into evidence that this formalized epistemological


method is quite fundamentally distinct from a computational reduc-
tion.

Furthermore, MRC excludes also the conceptual reductions:

Π17. Anti-reductionist proposition. Inside MRC the “reduction”


of a metadescription of order n (D.16) to the descriptions and elements of
descriptions of order n − k, k = 1, 2, . . . , n − 1 involved in it, is in general
impossible.
“Proof”. Consider the metaobject-entity œ(n) ) from a meta-
description which, inside the considered chain, is of order n, D(n) /G(n) , œ(n) , V (n) /.
(n−1)
An isolated element from œ(n) (a description Dj of order n − 1, or some
other descriptional element of order n − 1 from such a description (genera-
tor, object-entity, view)) in general simply does not exist in the sense of D7
with respect to the new meta-aspects of order n from V (n) . For instance, a
metaview V (2) of order 2 from the metadescription D(2) /G(2) , œ(2) , V (2) / rel-
atively to D(1) /G(1) , œ(1) , V (1) /, can contain the aspect of distance between
two space-gk-qualifications of order 1 involved by D(1) /G(1) , œ(1) , V (1) /,
170 M. MUGUR-SCHÄCHTER

with respect to which these qualifications themselves do not exist in the


sense of D7. Or else, œ(2) can contain two previously achieved descriptions
(1) (1)
of physical object-entities, DA and DB involving both a same view V (1)
(so qualifications of a same nature) while V (2) contains a meta-aspect of
(1)
order 2 of comparison of these qualifications, whereas neither DA alone nor
(1)
DA alone, nor descriptional elements from these, do exist in the sense of D7
with respect to this meta-aspect of comparison. In general terms now, the
new qualifications of order n that can be involved in a metadescription D(n)
while they cannot be involved in the descriptions of order n − 1 contained in
D(n) , consist of global or connective metaqualifications of order n concerning
two or more descriptional entities of order n − 1 from the object-entity œ(n)
from D(n) (consisting of whole descriptions of order n − 1, or generators
of object-entities, or object entities or views, of order n − 1). These, when
considered separately inside the descriptions of order n − 1, do not exist in
the sense of D7 with respect to any of such new metaqualifications of order
n involved by D(n) .
So in general D(n) is not reducible to the descriptions or descriptional
elements of orders n-k from the same chain.
The biologists should pay particular attention to this circumstance:

Comment. On each descriptional level of a given order n from a


descriptional chain (D.16), the descriptional cell D(n) placed on this level
introduces, via the condition of relative existence D7, the possibility of new
qualifications, of which the very definibility and meaningness are conditioned
by the previous achievement of the descriptions from all the previous levels
n − 1, n − 2, . . . , n − n:
Throughout the development of a process of conceptualization
normed accordingly to MRC one can literally watch the creative com-
plexifying work of cognitive time: One can literally see what “emer-
gence” means.
It is remarkable that inside MRC this conclusion follows from the
system of basic definitions, postulate and principles, in a way that permits
a clear perception of the nature of each contribution to the conclusion. One
can distinguish between contributions of a factual nature as for instance
those brought in by a basic description D(o) , and on the other hand con-
tributions of psychological nature like the choices of epistemic referentials
for the successive descriptional cells, or of methodological nature like the
condition D7 of mutual existence and the principle of separation P15:
There is no need any more for pleading, arguments, etc. in order to
draw attention upon the specific character, the mechanisms and the
QUANTUM MECHANICS VS. RELATIVIZED CONCEPTUALIZATION 171

features of what is labelled by the words “complexity”, “complexifi-


cation”, “emergence”.
So, by normed complexification, the transferred descriptions that
start from the inside of pure factuality and by which phenomena acquire
a first communicable form, are then developed in unlimited chains of hier-
archically connected metadescriptions of increasing order. These chains can
meet and interact variously at various levels and thus they weave indefinitely
compexifying and non predictable forms of communicable significance.
The consequences of the association between the principle of sepa-
ration and the concept of relativized metadescription, are innumerable and
always important. But in the absence of a normed descriptional structure to
which any description be referable, they cannot be systematically identified
and controlled.

4.2.4. Reference and minimality of the MRC-realism

At this stage of the elaboration of MRC it is already possible to entirely


elucidate a posteriori the a priori somewhat obscure features introduced by
the definition D4 of a generator of object-entity (the posited one-one relation
G-œG ) and by the realist postulate P3 (cf. note 8). We shall now achieve this
by a succession of three propositions. Thereby also the reflexive character
of MRC will gain new illustrations, while the formalized character of MRC
will become still clearer.

Π18. Propositions on reference and minimal realism

Π18.1. (On comparability, identity, and the relation G-œ œ G ). A


basic object-entity is inexistent in the sense of D7 with respect to any
“comparison-view”: such a view is a metaview with respect to which only
descriptions exist in the sense of D7, never basic object-entities.

“Proof”. What is not already pre-qualified cannot be compared.


Only two (or more) previously achieved descriptions D1 and D2 can be
compared, and only concerning some definite aspect-view or view with re-
spect to which these descriptions do both exist in the sense of D7. One can
for instance ask: are D1 and D2 identical or different with respect to this or
that gk-value of the aspect-view Vg ? If Vg is absent in one or in both consid-
ered descriptions, the question is meaningless because D1 and D2 constitute
together a meta-object-entity (D1 , D2 )(2) that does not exist in sense of D7
(2)
relatively to a metaview of g-comparison, say Vgc , so a fortiori a gk-identity
can be neither established nor refuted. If on the contrary both D1 and D2
(2)
do make use of Vg , then (D1 , D2 )(2) and Vgc do satisfy D7 and so one can
172 M. MUGUR-SCHÄCHTER

research whether, yes or no, D1 and D2 do possess some gk-identities. In this


example, I have brought into play a most simple metaview of comparison
with respect to only one aspect g. Nevertheless this view is already, quite
essentially, a metaview. One can form much richer metaviews of comparison.
But all are metaviews relative to definite views with respect to which only
previously achieved descriptions can exist in the sense of D7.
A basic object-entity—a bulk of pure a-conceptual factuality—is not
a previously achieved description. Therefore it cannot be compared, neither
to “itself” nor to something else.

Comment. So the whole stratum constituted by the very first prod-


ucts of the epistemic actions—the stratum of basic object-entities introduced
by basic generators—is not reachable by the concept of comparison and by
the qualifications derived from it, identity, difference, degree of similitude.
For basic object-entities these qualifications cannot be established by inves-
tigation, they can only by posited by method (like in the definition D4 of
a generator of object-entity). When a given basic operation G(o) of gener-
ation of object-entity is repeated, it simply is meaningless to ask whether
yes or not the object-entities œ(o) produced by this operation are all identi-
cal: this finally founds “deductively” inside MRC the impossibility to assign
a general meaning to the question whether yes or not the repetition of a
given operation G of generation of an object-entity œG , produces identical
results œG . So the posit of a one-one relation G-œG appears a posteriori to
be necessary indeed in order to be always able to speak and think fluently
concerning the products of G; while the significance of this posit, already
specified to a certain degree in the comment on π12, becomes now fully
clear.
The one-one relation G-œG founds a methodological strategy accord-
ing to which the referent œG —independently defined from the start, namely
by the operation G, and posited to be unique—associates coherently with,
both, the a priori condition of possibility in the sense of D7 of an as yet
non-defined meaning of œG with respect to a given view V , and with a sub-
sequently constructed specified meaning of œG with respect to V (while for
another view V 0 6= V , the relative existence D7, or a meaning of œG , or
both, might fail to exist).

Thus the question of reference obtains a self-consistent and effective


solution.

Π18.2.“Local” proposition on the realist postulate. Consider a


physical object-entity œG . This is a fragment of physical reality generated by
a given physical operation of generation G. The fact that any communicable
QUANTUM MECHANICS VS. RELATIVIZED CONCEPTUALIZATION 173

knowledge is description, and the relativity of any basic description to a


basic view, entail that the sequence of words “knowledge of how œG is in
itself” is void of significance.

“Proof”. Consider a physical object-entity œG . Any communicable


knowledge concerning œG amounts to some relative description D/G,œG ,V /.
Any relative description D/G,œG ,V / belongs to some net of descriptional
chains that is rooted in pure factuality via a (finite) number of basic trans-
(o)
ferred descriptions D(o) /G(o) , œ(o) , V (o) / the basic object-entities œG from
which somehow contributed to œG , have hereditarily transmitted into œG
some of their own semantic substance. Now, in each one of these basic trans-
ferred descriptions, the transfer-view V (o) acting there yields for the involved
(o)
basic object-entity œG a very first access to observability. But the princi-
ple P10 of individual mutual exclusion, the propositions π11, π12, π13, and
the definition D14.3.1 of a basic description, show that, and how, the basic
transfer-view V (o) , while it yields this first access, also inserts a non re-
movable opaque screen between the acting consciousness-functioning CF and
(o) (o)
“œG -in-itself”, it bars the way of human knowledge toward “œG -in-itself”.
So the unavoidable and non removable descriptional relativities explicated
inside MRC, and the fact that any communicable knowledge is description,
entail inside MRC that [knowledge-of-the-physical-reality-as-it-is-in-itself] is
nothing more than a meaningless combination of words, devoid of any des-
ignatum.

Comment. Since Kant the impossibility to know how a physical


entity “is-in-itself”, is accepted as an obvious postulate inside philosophy.
But many physicists still are reluctant to fully realize this definitive limit
of human rational knowledge. So it seems worth mentioning explicitly that
inside MRC this limit follows from the posited assumptions without being
one of these. So that there is no need to assert it as a logically independent
assumption. Then those who contest this limit should specify which posited
assumptions they contest.

Π18.3.“Global” proposition on the realist postulate: mini-


mality. Inside MRC the realist postulate P3 can only be given a minimal
significance: it can only be understood to assert exclusively the credo of
the existence, apart from the interior reality from my own mind, of also
a physical reality independent of any act of observation; but an existence
which is strictly non-qualifiable “in-itself”, beyond the mere trivial and non-
informative, idempotent assertion of its relativized qualifiability, if acts of
observation of it do take place in the conditions D4-D7 (in the absence of
174 M. MUGUR-SCHÄCHTER

which P3 would be aimless).

“Proof”. According to the definition D2 “the physical reality”, glob-


ally considered, is just a posited substratum wherefrom all the basic object-
(o)
entities œG considered in π18.1 and in the proof of π18.2, are conceived to
be extracted. Only this and nothing more. It would then be an arbitrary
conceptual discontinuity, a leap, a kind of spontaneous generation, of Deus
ex Machina, and even an inner inconsistency, to assign to this substratum
posited by us, properties that transcend the very descriptional essence of all
(o)
the fragments œG that we extract from it, namely the impossibility shown
by [π18.1+π18.2], to know any qualification whatever concerning a basic
(o)
object-entity œG in-itself.

Comment. It is quite non-trivial that inside MRC this minimality


of the realist postulate P3 is a feature that emerges as a consequence—in the
weak sense that marks all the “proofs”—of the non removable descriptional
relativities. So much more so that the forces which withstand the distinction
between mere existence of something, and knowledge of how this something
is, are huge.

Final global comment on the realist postulate (cf. note 8).


By now, I think, the specificity of the concept of “physical reality” with
respect to the general concept of reality introduced by D2, has come out with
satisfactory definiteness, mainly via the frame principle P8, the principle P10
of individual mutual exclusion, the propositions π11, π12, π13, the concept
D14.3.1 of basic transferred description, and the propositions from this point
18. Thereby, retroactively, the necessity of the postulate P3 as well as its
significance should have become clear. This necessity lies in the fact that the
formulations mentioned above would not have been possible without P3. As
for the significance of P3 inside MRC, it can be best grasped per a contrario:
it is that which inside MRC makes no sense, or no clear sense, when one
considers elements of reality consisting of concepts, social facts, etc.
As for the minimality of the realism asserted here, I suppose that
notwithstanding the proposition π18.3 many will tend to continue to nur-
ture in their minds a non-minimal realism. But reconsider in full light the
quasi irrepressible hope that, in spite of all, some model or “only some
invariants”, might some day transpierce the obstacle generated by the de-
scriptional relativities and inform us definitively, even if only in a coded
way, on how the physical reality is-in-itself, independently of any percep-
tion. And on the other hand, consider the necessarily fragmenting character
of the knowledge that human mind can construct, the indefinite and evolving
QUANTUM MECHANICS VS. RELATIVIZED CONCEPTUALIZATION 175

multiplicity of the possible basic object-entities œ(o) as well as of the basic


transfer-views V (o) which—now or in the future—could be found to exist in
the sense of D7 with respect to a given basic object-entity œ(o) : these stress
even more, if still possible, the illusory character of such a hope for non-
minimality. Indeed, given the non removable dependence of thought on per-
ception, given the non removable dependence of perception on fragmenting
descriptional relativities, given the unpredictable and incessant complexifi-
cations brought forth by the so various, and unbounded, hierarchical chains
of metadescriptions that are growing everywhere, given the unpredictable
changes of “viewpoint” (of epistemic referential) which these complexifica-
tions might bring forth—certainly radical from time to time—on what a
rational basis could one uphold the postulation of some convergence toward
a definite, definitive, terminal, absolute descriptional structure (supposing
that this succession of words were endowed with some meaning)? What a
sort of invariants, magically stabilized against all the changes brought forth
by the growth of thought, and magically freed of any descriptional relativ-
ity, could, thus stripped, nevertheless carry knowledge of the way of being
of physical reality in-itself, beyond the posit of its mere existence? When
knowledge is nothing else than qualifications via some view, of a somehow
delimited object-entity, so qualifications relative to some view and some gen-
erator of object-entity? Obviously one ends here up in a whirl of circularity.

4.2.5. Relative models versus minimal realism

But if any knowledge-of-how-physical-reality-is-in-itself, is indeed an illu-


sory self-contradicting concept, why do our minds so stubbornly keep to this
concept? This is a question which deserves being examined.
So I close now this exposition of the nucleus of MRC as follows. First I
shall show why the illusory belief in the possibility to reach knowledge of how
physical object-entities are in-themselves, is quasi irrepressibly generated
by human mind, in consequence of the frame-pronciple P8. And then I shall
show how, once identified, the fallacy vanishes and leaves place to dimensions
of conceptual liberty.
I proceed by defining a last group of four concepts which specify
entirely the philosophical status of the minimal realism asserted here.

On the insufficiency of the basic transferred descriptions.


Consider first an individual transferred description D(o) /G(o) ,œ(o) ,V (o) / of
(o)
a physical basic object-entity œ(o) (i.e., for any aspect-view Vg ∈ V (o) ,
(o)
when the succession [G(o) .Vg ] is repeated, always the same value gk is
obtained). In this case, by hypothesis, the epistemic referential (G(o) ,V (o) )
insures for the transferred results the strongest possible sort of qualificational
176 M. MUGUR-SCHÄCHTER

stability (π12, π13, D14.1). While furthermore, according to D14.3.1 the


basic transferred description D(o) characterizes observationally the involved
basic object-entity œ(o) . So one finds oneself already in possession of an
observational invariant that associates a quite definite meaning to what has
been labelled a priori “œ(o) ” (cf. the comments on the final posit from
D14.1.3). It might then be argued that this “suffices”, that in such conditions
there is no reason for researching further specifications concerning what
has been labelled œ(o) . But the fact is that in general such a “sufficiency”
simply is not experienced by the observer-conceptors: in presence of even an
individual transferred description D(o) that produces a most immediately
manifest observational stability, many thinkers (if not most)—quite modern
thinkers, and even physicists—experience an irrepressible tendency toward
a subsequent epistemic elaboration that shall produce a better, a clearer
meaning assignable to what has been labelled œ(o) . But a “better, clearer
meaning of oe”, in what a sense, exactly?
When one tries to answer this question it appears that what is re-
searched is a representation of œ(o) that shall endow it with an own form
of spacetime-gk-values, separated from any process of observation and any
registering device; and moreover a form of spacetime-gk-values possessing
“unity”, i.e., covering a connected space-domain obeying some definite dy-
namical law.
Furthermore a global and explicit spacetime representation is
(vaguely) desired for also the processes that have led from the basic object-
entity œ(o) with its own spacetime location, to its basic transferred descrip-
tion. The frame-principle P8 is here at work.
The requirements of the frame-principle cannot be violated defini-
tively. One can at most postpone dealing explicitly with them. The frame-
principle expresses a psychical fact which is as irrepressible as the physical
fact that masses are tied with gravitation. If a basic transferred description
of a basic object-entity is asserted, then one should be able to imagine some
possible own form of spacetime gk-values of this object-entity, as well as
some possible own structure of spacetime gk-values of the process that has
generated the description. If not, the frame principle will keep active and
upset us.
A basic transferred description D(o) , though, yields no hint for sat-
isfying these requirements. It is expressed exclusively in terms of observable
features of registering devices which are all distinct from what is la-
belled œ(o) . It yields no representation whatever concerning the spacetime
location of the basic object-entity œ(o) itself. Inside a basic description D(o)
the involved basic object-entity œ(o) is not represented as an autonomous in-
dividuality endowed with an own form, it still floats behind as a mere labelled
QUANTUM MECHANICS VS. RELATIVIZED CONCEPTUALIZATION 177

nebula suggested by the words basic object-entity and their notation œ(o) .
And even if, for a moment, we suspend any question concerning specifically
œ(o) and we consider D(o) as a whole, again we find ourselves in presence
of an absence of spacetime intelligibility. Indeed, given that each registered
mark gk involved by D(o) is found on a g-apparatus and that the transfer-
view V (o) must involve at least two different g-apparatuses for measuring
two mutually incompatible basic aspect-views, the “form” of spacetime gk-
values involved by the basic transferred description D(o) itself is found to
cover a scattered domain of space, tied with different registering devices that
can lie arbitrarily far from one another. And given that the time-origin to
(o)
has to be re-established after each realization of a succession [G(o) .Vg ], it
is not even clear whether it is possible to somehow associate this form with
some continuous evolution (or persistence) ordered by a unique increasing
time-parameter.
In short, by D(o) alone one cannot “understand” intuitively, neither
how the basic object-entity can be conceived to “be”, nor in what a sense,
exactly, D(o) is a “description” of this basic object-entity. This situation
is tiring for the mind. Therefore an individual basic transferred description
D(o) is not perceived as an achieved descriptional action. It is not felt to have
reached a conceptual stage of epistemological equilibrium. It is obscurely felt
as if loosely fixed on a steep conceptual slope where a conceptual force draws
it toward a separated representation of œ(o) in terms of own gk-spacetime
aspect-values. This sort of need might be regarded as a methodological in-
stinct tied with the frame-principle, induced by the adaptive biological evo-
lution of our minds.
All the preceding remarks hold also concerning a probabilistic trans-
ferred description. The now seventy years old debate on the interpretation
of quantum mechanics proves this enough.
So one is led to consider the following question: is it possible to elab-
orate, out of a previously achieved basic transferred description D(o) , a sep-
arated description of the basic object-entity œ(o) involved in D(o) ? Not a
description of “how œ(o) really is”—by now such naı̈ve epistemic quests can
be supposed to have been entirely transcended inside MRC—but a spec-
ification of just a possible modus of thinking of œ(o) in a self-consistent,
transparent, intellectually operational way that be naturally insertable into
the current language-and-conceptualization. The answer to this question is
positive and it is brought forth by the following three new definitions.

D19. Intrinsic metaconceptualization. Intrinsic model

D19.1. Intrinsic metaconceptualization of a basic transferred de-


178 M. MUGUR-SCHÄCHTER

scription. Consider a basic transferred description D(o) of a physical object-


entity œ(o) , individual or probabilistic.

- Let G(1) be a metagenerator of object-entity consisting of a concep-


tual selector (D4) that selects for examination the meta-object-entity
consisting of œ(1) ≡ [D(o) + œ(o) ].
(1)
- Let VI indicate an intrinsizing metaview (I: intrinsizing) which,
starting from the initial, purely observational, transferred description
D(o) , works out intrinsic qualifications of the basic object-entity œ(o)
involved in D(o) (intrinsic: word used in order to distinguish from the
philosophical term “in itself”). This, inside the new epistemic referen-
(1)
tial (G(1) ,VI ), is achieved as follows:

(1)
* Let VIg (I fixed, g = 1, 2, . . . , m, Ig functioning as one com-
pact index) be a set of m intrinsizing meta-aspect-views which,
(1)
together, constitute the intrinsizing metaview VI .
(1)
* Each intrinsizing meta-aspect-view VIg involves an abstract, con-
(1)
ceptual VIg -operation of examination of the metaobject-entity
[D(o) + œ(o) ], namely an examination constructed in a way such
(1)
that its possible results – necessarily values (Ig)k of VIg , ac-
cordingly to the definition D.5.1—are all conceivable as separate
intrinsic qualifications (Ig)k of the basic object-entity œ(o) , which
are compatible with D(o) .
(1)
* The values (Ig)k of the intrinsizing metaview VIg are further-
more constructed as: (a) intrinsic qualifications of œ(o) at the
time to which is the time-origin re-established at the beginning
of each succession [G.Vg ] having contributed to the elaboration of
D(o) ; (b) qualifications located inside a connected space-volume
∂r (r: space-position vector inside the spacetime referential in-
volved by D(o) according to P8 and C9) which œ(o) is posited to
occupy at the time to .

(1)
The relative metadescription D(1) /G(1) , œ(1) , VI / constructed as
specified above will be called an intrinsic metaconceptualization of the basic
(individual or probabilistic) transferred description D(o) /G(o) , œ(o) , V (o) /
and it will be also assigned the alternative more specific symbol
(1) (1)
DI /[D(o) , VI ].

Comment. We speak of “an” (not “the”) intrinsic metaconceptual-


ization of D(o) , because in general many different intrinsizing metaviews can
QUANTUM MECHANICS VS. RELATIVIZED CONCEPTUALIZATION 179

be constructed, and each one of these yields a corresponding and possibly


specific intrinsic metaconceptualization.
An intrinsic metaconceptualization of a basic transferred description
D(o) realizes a retro-active localizing projection of the scattered form of D(o) ,
onto a connected and instantaneous spacetime domain [∂r to ]. The unique-
ness of the temporal qualification to , even though it is retro-active, suf-
fices now for permitting to posit, starting from it, an intrinsic time-order
that is hidden to observation. This permits now to assign a law of intrinsic
evolution to what has been labelled œ(o) , underlying any evolution of the
observable transferred description D(o) . As for the transferred description
D(o) , it can now finally be explained. The basic object-entity œ(o) can now
be conceived to have “possessed” at the time to —on the connected spatial
domain ∂r —the features assigned to it by the intrinsic metaconceptualiza-
(1) (1)
tion DI /[D(o) , VI ]. These, one can now think, were own features of œ(o) ,
separated from those of any measurement device, independent of them, but
features which D(o) has been able to transpose into observable manifesta-
tions, only by disorganising the form of intrinsic gk-spacetime aspect-values
constituted by them. The scattered form of spacetime-gk-values involved by
D(o) can now be thought of as the result of a bursting and change of the
initially integrated intrinsic features of œ(o) itself. A bursting produced by
(o)
the mutual incompatibility of certain aspect-views Vg from the transfer-
view V (o) which has obliged us to perform a set of different successions
(o) (o)
[G(o) .Vg ], Vg ∈ V (o) in order to obtain the global transferred description
(o)
D(o) (according to D14.1 at least two such incompatible aspect-views Vg
are necessary in order to characterize œ(o) ).
In short, by the assumptions from D.19.1 the basic object-entity œ(o)
has acquired the specification of an own form of gk-spacetime aspect-values,
and the process of emergence of the basic, transferred description D(o) has
been causalized: the categories of space, time and form have been restored
for œ(o) , so D(o) has now become intelligible. But thereby the frontier be-
tween the two strata of our conceptualizations—the primoridal stratum of
basic transferred descriptions D(o) which draws in fragments of pure fac-
tuality, and the second stratum consisting of a web of indefinite chains
of metadescriptions D(n) , n = 1, 2, . . . (see the comment on D16)—is now
crossed: from now on we find ourselves inside the unlimited depth of verbal-
symbolic developments of the basic transferred descriptions.

D19.2. Intrinsic model of a physical basic object-entity. So


(1)
the intrinsic metaconceptualization DI /[D(o) , VI (1)] constructs “explana-
tory” relations between its global meta-object-entity œ(1) ≡ [D(o) +œ(o) ] and
the basic object-entity œ(o) involved by D(o) , as well as an own spacetime
180 M. MUGUR-SCHÄCHTER

representation of this basic object-entity œ(o) . Once this construction has


been achieved it is possible to extract from it exclusively the representation
of the basic object-entity œ(o) , in the following way.
The set of intrinsic qualifications of the basic object-entity œ(o) pro-
(1)
duced by the intrinsic metaconceptualization DI /[D(o) , VI (1)], when con-
sidered alone, severed from all the other elements with which it is tied inside
(1) (1)
the intrinsizing metadescription [DI /D(o) , VI ], will be called an (intrin-
(1)
sic) model of œ(o) and will be symbolized by M (œ(o) )/[V (o) , VI ] in order
to remind explicitly of the non-removable relativity of this model to the pair
(1)
of views [V (o) , VI ] which determined its genesis and its characters.

Comment. It is important to realize clearly that an intrinsic model


(1)
M (œ(o) )/[V (o) , VI ]
is not a relative description of œ(o) in the sense of the
definitions D14.
(1)
The intrinsizing meta-aspect-views from VI that produced the qual-
(1)
ifications assigned to œ(o) by the intrinsic model M (œ(o) )/[V (o) , VI ],
have examined the meta-object-entity œ(1) ≡ [D(o) + œ(o) ], not the
basic object-entity œ(o) .
(1)
The model M (œ(o) )/[V (o) , VI ] occupies finally a position of full sat-
uration and equilibrium of the meaning assigned to what had been initially
labelled œ(o) . Its genetic compatibility with the transferred description D(o) ,
(1) (1)
as represented by the intrinsizing metaconceptualization [DI /D(o) , VI ],
detached it from D(o) like a mature fruit that has been plucked from its
(1)
tree. The model M (œ(o) )/[V (o) , VI ] superposes now to the initial purely
observational basic description D(o) , a pragmatic, economic and stable con-
ceptual closure. Namely a closure consisting of an invariant with respect to
(o) (o)
the group of transformations from one succession [G(o) .Vg ], Vg ∈ V (o) ,
(o)
which contributed to the elaboration of D , to any other such succession
with a different aspect-view in it, G(o) being fixed : the observable effects
(o) (o)
of all these different successions [G(o) .Vg ], Vg ∈ V (o) , are now all assigned
(1)
one common and definite “causal” ancestor M (œ(o) )/[V (o) , VI ] which pro-
duces various perceptible manifestations, in a “normal” way, which means
in a way that is understandable accordingly to the frame-principle P8.
When the basic transferred description D(o) on which the model
(1)
M (œ )/[V (o) , VI ] is founded involves exclusively the human biological
(o)

sensorial apparatuses, this sort of closure emerges in an unconscious, non-


mediated, genetically wired way: It is precisely what we believe to per-
ceive, and this we automatically assign to, exclusively, the involved object-
entity....in-itself . The stage of a transferred description D(o) remains un-
QUANTUM MECHANICS VS. RELATIVIZED CONCEPTUALIZATION 181

known and unsuspected. And even when fabricated apparatuses are con-
nected to the biological ones, if the whole apparatus thus obtained still
offers a directly intelligible form of spacetime gk-values, this form, again, is
irrepressibly felt to reveal how the perceived object-entity is in-itself (think
of perceptions via a microscope or a telescope). Moreover, when, as in quan-
tum mechanics, the observable basic transferred data do not themselves offer
a directly intelligible form of spacetime gk-values, so if an intrinsic model
(1)
M (œ(o) )/[V (o) , VI ] has to be explicitly constructed from these data treated
as mere coding signs, still, once a model has been constructed, it usually is
felt to be satisfactory and necessary to such a degree that its only hypo-
thetical, retro-active, and relative character tends to be skipped. Implicitly
(1)
and fallaciously the intrinsic models M (œ(o) )/[V (o) , VI ] conquer inside our
minds a primary and absolute status.
This is the fallacy that instates the irrepressible belief that physical
object-entities can be known “such as they are in themselves”.
The unavoidable dependence of any intrinsic model of œ(o) , on both an initial
transferred description D(o) that has had to be achieved first and has in-
volved some particular transfer-view V (o) , and a subsequent process of intrin-
(1)
sic metaconceptualization DI involving a particular intrinsizing metaview
(1)
VI , tends to be overlooked. In particular, it tends to remain unnoticed that
(1)
another pair (V (o) , VI ) would have led to a different model of œ(o) .
These occultations mark all the classical descriptions, in physics, in
mathematics, etc., as well as in the current thinking expressed by the
current language: they are the opaque fictitious platform that floats
above the physical factuality and on which is erected the classical
concept of objectivity. The roots which insert the conceptualizations
into physical factuality, with the relativities involved by them, are
hidden beneath this fictitious platform.
Starting from the transferred data that are available for it and on
which it takes support without trying to express them, human mind always
rushes as rapidly and as directly as it can toward a representation of the
involved object-entity by an intrinsic spacetime-gk-values model. As soon as
such a representation has been attained, it is spontaneously felt to be “true”
in a primary, certain and absolute way, without reference to the initial trans-
ferred data on which it is founded and forgetting that it is just an economic,
hypothetical, retro-actively imagined construct. While the initial transferred
data, even though they are the sole certainties, because of their dispersed
unintelligible phenomenal appearance, are implicitly and irrepressibly per-
ceived as nothing more than “subjective” tools for finding access to the
182 M. MUGUR-SCHÄCHTER

“objective truth”:a fallacious, illusive inversion. We systematically commit


what Firth [20] called “the fallacy of conceptual retrojection”. Simplicity,
invariance, and what we tend to call “truth” and “objectivity”, have coa-
lesced in a knot imprinted upon our minds by ancestral processes which, by
implicit pragmatic causalisations, optimizes the efficiency of our behaviour,
but blocks and botches the reflexive knowledge of our fundamental episte-
mological functioning. The interpretation as ontological assignments, of the
results of our instinctive human adaptive constructs involving the frame-
principle, is one of the worst and most stubborn pathologies of thought.
But in quantum mechanics this process has hit an obstacle. Up to
(1)
this very day a type of intrinsic model M (œ(o) )/[V (o) , VI ] fitting satisfac-
torily the quantum mechanical transferred descriptions of what is called a
microstate, has not yet been found. So it has been necessary to stop the
attention upon these transferred descriptions themselves such as they have
emerged, and to embody these transferred descriptions in mathematical ex-
pressions able to yield, if not understanding, at least numerical predictions.
And then, like a tireless insect when its instinctive constructive actions are
hindered, human mind came back again and again upon these quantum
mechanical transferred descriptions that resist modelization. And so it has
become possible to discern more and more explicitly their specificity, which
inside MRC has been redefined in quite general terms and has been called
a “basic transferred” character. In this way we finally become aware of the
unavoidable necessity of a quite universal first phase of conceptualization in
terms of basic transferred descriptions.

Inside MRC the distinction between illusory ontological assertions


concerning an absolute way in which œ(o) “really-is-in-itself”, and
relative methodological intrinsic models of œ(o) , is quite radical, elab-
orate and clear cut. And the genetic order of the descriptional steps
is re-constructed correctly and is fully displayed.

Under these conditions the irreplaceable pragmatic and heuristic


power of intrinsic models can be put to work without triggering any more in-
soluble philosophical pseudo-problems. Correlatively, the vain and exhaust-
ing battle between positivists and defenders of modelization, evaporates.
The transferred descriptions are the unavoidable first stage of our processes
of conceptualization, while the intrinsic metaconceptualizations of the initial
transferred descriptions and the relative models extracted from these are a
stabilising subsequent stage which, if realized, brings us down onto a (local,
psychological and provisional) minimum of our potential of conceptualiza-
tion.
QUANTUM MECHANICS VS. RELATIVIZED CONCEPTUALIZATION 183

There is no choice to be made. There is just an unavoidable order


of elaboration to be observed, in a normed way, or to be recognized
when it occurs implicitly.

D19.3. Minimal intrinsic metaconceptualization. Minimal


intrinsic model. Consider a basic transferred description D(o) of a phys-
ical basic object-entity. The effect labelled œ(o) of the basic operation G(o)
of generation of an object-entity can always be trivially metaconstructed
accordingly to D19.1 so as to be conceivable as:

A bulk of potentialities of future observable manifestations, deter-


mined by G(o) on a finite space-domain ∂r3 , at the time to when G(o)
comes to an end, each one of these potentialities being relative to an
(o)
aspect-view Vg from the basic view V (o) operating in D(o) .

For this it suffices to posit in D19.1 the minimal intrinsizing view corre-
(1)
sponding to V (o) —let us denote it [min.VI /V (o) ]—defined as follows. For
(o) (1)
each basic aspect-view Vg from the basic view V (o) , [min.VI ] contains a
(1)
corresponding intrinsizing minimal meta-aspect-view [min.VIg ] possessing
a unique minimal meta-aspect-value denoted Igmin that consists of the in-
trinsic potentiality, assigned to what has been labelled œ(o) , to produce at a
time tg > to , any one among the transferred observable aspect-values gk of
(o) (o)
the basic aspect-view Vg , iff œ(o) is subjected at to to an Vg -examination
(o)
(tg − to : the duration of a Vg -examination, characteristic of the considered
aspect g) (I recall that “intrinsic” means here assigned to œ(o) itself as an
own feature, the word having been chosen in order to distinguish from the
meaning of the philosophical term “in itself”).
The trivial realization of the definition D19.1 specified above will be
called the minimal intrinsic metaconceptualization of the basic transferred
(1)
description D(o) /G(o) ,œ(o) ,V (o) / and it will be denoted by [min.DI /D(o) ]
(1)
(the relativity to the acting intrinsizing view VI is now included in the
definition of the minimal intrinsizing view [min.VI (1) /V (o) ], so it is ab-
sorbed in the proper “min.”). The intrinsic model of œ(o) extracted from
(1)
[min.DI /D(o) ] will be called the minimal intrinsic model of œ(o) and will
be denoted by [min.M (œ(o) /V (o) ].
Comment. The following consequence of the final posit from
D14.3.1 is quite worth being noticed. Any basic view V (o) that involves
(o) (o) (o)
two mutually incompatible basic aspect-views Vg1 and Vg2 6= Vg1 entails
a minimal intrinsic model [min.M (œ(o) /V (o) ] which now characterizes œ(o)
conceptually (by predication). It yields a conceptual definition of œ(o) that
can now be added to the purely factual definition of œ(o) insured initially
184 M. MUGUR-SCHÄCHTER

by the operation G(o) alone (whereby œ(o) still remained outside knowl-
edge) and to the subsequent purely observational description of œ(o) offered
by the basic description D(o) (whereby œ(o) , though characterized obser-
vationally, nevertheless was still devoid of an own conceptual representa-
tion). MRC brings forth degrees of characterization of a basic object-entity
œ(o) , which compose the complexifying sequence [purely factual→ purely
observational→ conceptual]. From that stage on, chains of non minimal in-
trinsic metaconceptualizations can indefinitely increase the degree of concep-
tual characterization of œ(o) . This illustrates the reflexivity of the method
and its unlimited character.
As any intrinsic metaconceptualization and any intrinsic model, the
trivial minimal models also may be perceived as “opportunistic” constructs
where what is actually observed is posited to stem from an a posteriori
imagined ad hoc explanatory potentiality. This however does not in the least
diminish the pragmatic importance of the fact that a minimal model of what
is labelled œ(o) is a representation that permits a most natural, easy insertion
of œ(o) into the conceptualization. Moreover it is always and automatically
realizable. It is however useful to remember again and again that inside
MRC this sort of representation is accepted as just an unavoidable strategic
step that must be carefully distinguished from an ontological credo: nothing
whatever is naively asserted concerning the impossible question of how the
basic object-entity œ(o) “really-is-in-itself”. It is only stated how this object-
entity can be most simply conceived in order for us to become able to speak
and think of it in structured, consistent, fluent terms.

4.2.6. Final comment on the realism involved in MRC

“ . . . Thus the aim of the book is to draw a limit to thought – not


to thought, but to the expression of thoughts: for in order to be able
to draw a limit to thought, we should have to find both sides of the
limit thinkable (i.e., we should have to be able to think what cannot
be thought).”
—Ludwig Wittgenstein, in the Preface to his Tractatus.

The concept of minimal realism possesses, I think, an essential philo-


sophical importance. Imagine an abstract surface on which are displayed all
the grammatically correct structures of words that human mind can compose
about the physical reality. On this surface, the concept of minimal realism
is delimited by a boundary which coincides strictly with the boundary that
separates the domain of communicable knowledges, from the domain inside
which can be found only expressions that are grammatically correct, so com-
municable, but devoid of reference: This boundary defines the extreme limit
QUANTUM MECHANICS VS. RELATIVIZED CONCEPTUALIZATION 185

that expressions of communicable knowledge can reach. The communicable


knowledges cannot transcend this frontier. They can just advance toward
it and eventually hit it by this or that basic transferred description which
acts like a small squad carrying a local net of pre-conceptualization inside
which it captures a small load of as yet unknown physical factuality which
it hoists up on the very first level of speakable, communicable knowledge.
But thereby the progression of the squad from inside the zone of knowledge,
toward the physical reality, is stopped. The squad is reflected back like an
elastic ball toward the inside of the realm of relative descriptions, where it
delivers its load which, from that moment on, can indefinitely be elaborated
along innumerable branches of complexification by intrinsic metaconceptu-
alizations and/or by extraction of corresonding intrinsic models. But each
one of these complexifying elaborations introduces new descriptional rela-
tivities which thicken the screen between physical reality in-itself, and our
mind’s representations of it, they thicken this screen so as to improve intelli-
gibility and thereby the capacity to think and to act. Such is the paradoxical
relation between physical reality and mind.
It is crucial to become aware, intensely, of the surreptitious advent
of this inversion in our direction of conceptualization, of these unavoidable
rebounds in the opposite direction each time that the extreme frontier of
the domain of communicable knowledge is hit by a basic description. If
not, we remain imprisoned in the inertial illusion that by modelizing more
and more we approach more and more the knowledge of how the physi-
cal reality “is-in-itself”. The grammatically correct associations of words
which express this illusion are founded upon a self-contradicting concept of
reality-in-itself, namely the concept of a qualifiable reality-in-itself. Whereas
reality-in-itself—by definition— is precisely what cannot be qualified more
than by its mere quali-fiability. By these words, “in-itself”, what is pointed
toward deliberately is nothing more than a posited existence, posited also to
be qualifi-able but to be devoid of any other more specifying qualifi-cation.

This is not a matter of fact; it is a matter of organization of


language-and-concepts.

The words “description” and “physical reality in itself” have to be


somehow endowed with a definition (even if only implicitly). And when
this is done in a coherent way what is called description is opposed by
construction to what is called “physical reality in itself”.
One might perhaps believe, for instance, that it is possible to gain one
more inch by specifying that the reality-in-itself is “such” that the qualifica-
tions which it admits from our part are precisely those which are elaborated
by our senses and our investigations. But when we focus attention upon
186 M. MUGUR-SCHÄCHTER

this supposedly supplementary specification, trying to capture an element


of positive novelty added by it to the minimal realist postulate, we find only
nothingness. We find ourselves placed on exactly the same content of infor-
mation as before. Any qualification added to that of mere qualifiable but
unqualified existence, even the most feeble one, the most vague, is either
idempotent, or generates contradiction. Any attempt to superpose some nu-
ance expressible in terms of approximations or of asymptotic apprehension
of how the physical reality is in itself, would only manifest a misunderstand-
ing of the nature of what is here involved, namely an optimized organization
of concepts-and-words. One can reasonably try to fight against a physical
circumstance, even if it is a “physical law”, trying to master it so as to realize
some technical aim. But trying to fight against the limitations entailed by a
conceptual-linguistic organization, manifests a confusion concerning essence:
what meaning would that have, for instance, to fight against the limitations
imposed by a previously constructed formal system, say arithmetic, which
one does not criticize and inside which one has placed oneself? “The-way-
of-existing-of-reality-in-itself” is a self-contradicting notion stemming from
a confusion between empirical circumstances and conceptual organizations
of which on the other hand one makes use.
In his Conference on Ethics, Wittgenstein said (concerning the more
or less similar confusion between value and truth): “It is perfectly, abso-
lutely hopeless to thus bump our forehead against the walls of our cage”
(my own retro-translation in English, from French translation). One can ap-
ply the same assertion to the confusion between an impossible ontological
quest, and an organization of language-and-concepts constructed by man.
This confusion entails chimerical aims and fictitious problems, like in the
quantum mechanical orthodoxy the arbitrary positivistic interdiction of in-
trinsic metaconceptualizations and intrinsic models because these are con-
founded with impossible qualifications of reality-in-itself. This mythic fauna
that spouts from the bursting of an inertially oriented impetus to understand
more, against the barrier placed by thought itself between communicable
knowledge, and a just posited and denominated rest, must be exorcised.
So the minimal realism involved by MRC has a composite logical sta-
tus. While the feature of minimality follows “deductively” inside the method
(π18), the main term, realism itself, is just a posit, the postulate P3. It is a
declaration of metaphysical belief, wholly subjective. Any question of truth
or objectivity is meaningless concerning it. But this metaphysical belief plays
a fundamental role for MRC: It seats the method on a unifying ground. It
asserts that beneath the endless proliferation of branching relativities which
mark the contents of descriptions, there exists a substratum of non referred
absolute, wherefrom the relativities emerge together with the conceptual-
QUANTUM MECHANICS VS. RELATIVIZED CONCEPTUALIZATION 187

izations. I say “beneath” in order to stress that the thesis of realism draws
out of the domain of language and descriptions. By the mysterious powers
of self-transcendence of language, this thesis acts like a verbal directional
indicator, pointed from inside the volume of the expressed, but which points
toward an existence from outside this volume. It grasps the attention, dis-
places it, and installs it at the very core of the non expressible. There, inside
this background of unverbalisable which it succeeds to designate, the re-
alist thesis fixes the ends of the threads with the help of which the basic
transferred descriptions web to one another the two regions that stretch out
on the two sides of the ghostly but insuperable wall between what is by
construction devoid of legal communicable expression, and the legally for-
mulated and communicable. In spite of the fact that we cannot “find both
sides of the limit thinkable”. This is the fundamental, the huge epistemolog-
ical innovation hidden in the quantum mechanical formalism, which inside
MRC is explicated, generalized, and organized in detail by radical and sys-
tematic relativizations. Whereby all the false absolutes are suppressed, not
only those which vitiate esthetics, ethics and metaphysics; for everywhere
thought is invested by hosts of false absolutes that generate pathological
tissues of illusory problems and paradoxes that blur out the sound limit
between the thinkable and mere nonsense.
It might seem that this background of non referred, because it is
absolute, is incompatible with the method of relativized conceptualization.
But, and it is important to stress this, MRC by no means banishes any
absolute. It banishes exclusively the false absolutes, those which hide de-
scriptional relativities of which the presence can be identified, and which, if
ignored, can generate illusory problems. But it is clear that when one con-
structs, it is unavoidable to posit certain absolutes. All the definitions from
MRC, principles, etc., as such, have nothing relative about them. They are
absolutes of the method, by the help of which the descriptional relativities
are defined. And the existence of a physical reality posited in P3 is also
an absolute of the method. This concept is introduced as just the absolute
reference without which thought would get lost in an unexplained profusion
of diversity; an absolute reference which unifies in one closed coherent whole
all the indefinitely evolving descriptional relativities defined by the method.
I confess that the beauty which, to my eyes, emanates from this
unification, appears to me irrepressibly as a sign of pertinence. Man and
“reality” form a whole, and the feeling of beauty that can emerge in a
human mind, intimately tied with coherence, has for me the significance
of an announcement that certain slopes of the real have been embodied
without having been violated. Whatever the unimaginable designatum of
the succession of words which I just aligned, I want to align them, for we
188 M. MUGUR-SCHÄCHTER

must somehow speak in order to communicate—paradoxically and in spite


of all—concerning the unspeakable.

4.2.7. Global remarks on the nucleus of MRC

MRC is:

* Explicitly founded upon the functioning of human mind, with


its cognitive aims. The choices of the epistemic referentials that gen-
erate the relativized descriptions, stem from the consciousness func-
tioning of the acting observer-conceptor. Each such choice expresses
a curiosity, a descriptional aim of this consciousness functioning. The
descriptional aims expressed by the successive choices of an epistemic
referential, inside a chain of conceptualization, mirror the evolution
of the descriptional aims of the acting consciousness functioning, and
thereby they determine the “direction of conceptualization”, step by
step. Inside MRC, in its present stage at least, the descriptional aims
do not follow from methodological prescriptions. This means the fol-
lowing:
No AI-machine could, by applying MRC, work like a human being,
without being directed by a human being. But an AI-machine endowed
with an “MRC-program” (if this were possible) and drawn by a man,
would work exactly like that man.
This specifies the difference between AI and MRC as well as the par-
ticularity of an “MRC-program”.
* Explicitly rooted in pure factuality, which entails the possibil-
ity of a systematic and constructed distinction between potentiality
of an infinity of processes of actualization of relative observable man-
ifestations, and this or that actualized observable manifestation (cf.
the concept of “genset” in Sec. 5.2.2). Thereby it brings in the modal
dimension potential-actualization-actualized.
* Radically relativizing. The whole approach bears the seal of the
relative mutual existence of object-entities and views (or, equivalently,
of generators of object-entity, and views) and of the relativities of
descriptions to the triads G,œG ,V .
* Methodological, normative, legalizing. MRC is not an attempt
at describing the natural processes of conceptualization. Though data
(introspective, linguistic, etc.) concerning these natural processes are
strongly taken into account, nevertheless MRC recognizes the impossi-
bility of a “purely” descriptive account on the processes of description.
So, deliberately, it takes distance with respect to such an aim, by
constructing definitions and principles conceived in order to optimize
QUANTUM MECHANICS VS. RELATIVIZED CONCEPTUALIZATION 189

the processes of conceptualization in compatibility with definite goals,


namely the a priori elimination of any false absolutization, reflexiv-
ity, construction of a conceptual structure with respect to which it be
possible to “localize” any other descriptional structure, natural or not,
etc. Thereby MRC is formalized. Not yet mathematically and quan-
titatively formalized, like a physical theory, but already formalized,
qualitatively formalized.
* Finitistic, cellular, local. The fact that the construction of knowl-
edge requires parcellings, steps, is taken into account quite fundamen-
tally throughout MRC, via the principle of separation P15 and the
concept D16 of relative metadescription.
* Globally unlimited. Though everywhere there are strict local delim-
itations of the descriptional quest, which withstand any gliding into
relativism, globally nowhere a boundary is pre-imposed: the finalized
finitism of MRC generates infinities.
* Hierarchical. MRC generates hierarchical trajectories of conceptu-
alization, in contradistinction to the theory of logical types, or that of
levels of language, which introduce extended hierarchical strata.
* Directional and reflexive, endowed with a capacity for an a priori-a
posteriori double way progression. Before starting a given descriptional
cell, a free choice of the direction of conceptualization desired by the
observer-conceptor is expressed in a corresponding choice of an epis-
temic referential. Later the results of this choice can be rejected or
kept and developed, on the basis of explicitly defined criteria.

The various features enumerated above are not exhaustive. Nor, by no


means, are they mutually independent. Quite on the contrary, they all stem
from one core-structure that induces an innumerable host of connections be-
tween these features. This core-structure is dominated by the systematically
recurrent role of the consciousness-functioning which introduces the epis-
temic referentials. And along the whole hierarchy of distinct descriptional
cells of increasing order from each chain of conceptualization from the web
of such chains, the same fundamental MRC-requirements for a relativized
normed conceptualization manifest themselves with a sort of fractality: Each
time that an epistemic referential has been chosen—no matter on which level
of conceptualization—the generator of object-entity, the object-entity and
the view from it entail non removable descriptional relativities to them.

4.2.8. On the conceptual status of MRC.

To what class of conceptual beings does MRC belong? Any representation


of “natural facts” is more or less normative, never purely descriptional as
190 M. MUGUR-SCHÄCHTER

the classical myth of objectivity involves.

In the case of MRC the explicitly and resolutely methodological char-


acter is a major feature of the approach. Any confusion between on-
tological assertions or implications, and methodological constructs, is
most carefully avoided.

Nevertheless MRC can also be regarded as:

An attempt at a finitistic representation of the natural processes of


generation of meaning where both relativism and false absolutizations
are excluded ab initio by explicit rooting into pure factuality and by
deliberate systematic relativizations.

The fact that throughout the process of constructing MRC one acts
“logically”, is neither a circularity, nor does it involve that MRC is reducible
to a logic. It only illustrates the general reflexive, (a priori)-(a posteriori)
character of any approach and in particular of this one: a priori the logi-
cal criteria are supposed to be fulfilled and they are utilized implicitly [21],
but later, at a convenient level of development of the approach, the log-
ical criteria—as it will be shown in 5.1— become a posteriori explicitly
expressible in MRC-terms. (This sort of inner evolution partakes of the gen-
eral reflexive character of MRC that has permitted to admit a priori the
possibility of any pairing (G, V ) and to introduce only a posteriori criteria
concerning the relevance of a given pairing (G, V ): first became expressible
the criterion of mutual existence D7, and then the subsequent criterion of
stability involved in the definition D14.1).
So probably the best characterization is as follows.

MRC is a strongly normative representation of the processes of con-


ceptualization, of which the major specificities are: the place explic-
itly reserved to the consciousness functioning; the radical descrip-
tional relativizations; and the fact that it explicates the structure of
the very first step in the construction of objectivity, in the course
of which intakes of a-[conceptual-linguistic] fragments of pure factu-
ality adduce into language and thought the hard core of scientific
objectivity.

4.3. The Second Stage: an Ideographical Symbolization of


MRC
In all the expositions of MRC that preceded the present one, I included
in a presentation made in usual language, an ideographic symbolization
QUANTUM MECHANICS VS. RELATIVIZED CONCEPTUALIZATION 191

which—without being neither a formalization stricto sensu nor a mathemati-


cal representation—permits certain suggestive and economic expressions. In
this work I present it simplified and separately. In this way the symbol-
izations are made available while the drawbacks as well as the advantages
appear clearly:

- A consciousness functioning CF is represented by the sign  → suggest-



ing the whirling place from D1 that acts on both the Exterior Universe
and the Interior Universe where it belongs, and in particular also on
itself.
- Reality is again symbolized by the letter R.
- A generator G of object-entity will be represented by the sign ∆ and
will be re-named a delimitator of object-entity, in order to stress that,
whatever the nature of G, the final result is a delimitation, out of R, of
a corresponding object-entity. Thereby however one looses the accent
placed by the term “generator” upon a (possibly) radically creative
character of an operation of object-entity generation. Then:
- The “place” from R where ∆ works will be denoted R∆ .
- The object-entity produced by ∆ will be denoted by œ∆ .
- The process of delimitation by ∆, of an object-entity œ∆ , will be rep-
resented indifferently by

∆R∆ ⇒ œ∆ , or œ∆ ⇐ ∆R∆ ,

where the arrows do not have a logical meaning and cannot be considered
separately, they are cemented into the global symbolizations which read
respectively: “the delimitator ∆, acting on R at the place R∆ , produces the
object-entity œ∆ ”, and “the object-entity œ∆ produced by the delimitator
∆ that acts on R at the place R∆ ”. Notice that the introduction of these
symbolization permits us to distinguish between:

* ∆: an epistemic operator (in the sense of usual language, not of math-


ematics);
* ∆R∆ ⇒ œ∆ : a process, that mentions its beginning and its result;
* œ∆ ⇐ ∆R∆ : an explicit specification of an object-entity via the pro-
cess that produced it, which permits to specify an unobservable object-
entity, by the way of producing it.

Thereby the expressivity concerning this zone from MRC is considerably


increased.

- An aspect-view will be symbolized by the same sign Vg as before;


192 M. MUGUR-SCHÄCHTER

- The operation of examination of œ∆ by Vg will be represented by


Vg œ∆ .
Notice that the introduction of these symbols permits us to distinguish be-
tween:
* the epistemic operator Vg (in the sense of usual language, not of math-
ematics) and
* the operation of examination Vg œ∆ ,
which again is an increase of expressivity.
- A view will be symbolized as before by V .
- The global operation of examination of œ∆ , by V (achieved accordingly
to π11), will be represented by V œ∆ .
The remarks concerning Vg hold also concerning V .
- An epistemic referential is represented by (∆, V ).
- The representation of an observer-conceptor [CF, (G, V )] becomes
→, (∆, V )].
[↓
- The mutual inexistence between an object-entity œ∆ and a view V will
be symbolized by
6 ∃œ∆ /V or 6 ∃V /œ∆ ,
which reads, respectively, “the object-entity œ∆ does not exist with
respect to the view V ”, “the view V does not exist with respect to the
object-entity œ∆ ”.
- The mutual existence between an object-entity œ∆ and a view V will
be represented by
∃œ∆ /V or ∃V /œ∆ ,
which reads “the object-entity œ∆ does exist with respect to the view
V ”, “the view V does exist with respect to the object-entity œ∆ ”.
(All these symbolizations can also be used, in particular, with the
symbol of an aspect-view Vg instead of V , which changes the meaning
correspondingly).
- A spacetime view is represented as before by VET .
- The frame-principle can be symbolized in the following way:

[∃œ∆ /Vg ] → [∃VET : ∃œ∆ /(VET ∪ Vg )]

[6 ∃œ∆ /VET ], ∀VET , ∀œ∆ .


Here the arrow, quite independently of any connotation suggesting
formal logic, reads “entails that” (in the sense of natural logic); ∃ and
QUANTUM MECHANICS VS. RELATIVIZED CONCEPTUALIZATION 193

6 ∃ —outside any formal system, just in the sense of usual language or


of “natural logic”—read, respectively, “there exists” and “there does
not exist”; Vg denotes a physical aspect-view; VET ∪ Vg considered as
a one-block symbol, reads “the view formed with a spacetime view
VET and another physical aspect-view Vg ”. The global reading of this
symbolic picture is the verbal formulation of P8.
- The symbol of a relative description D/G,œG ,V / becomes
D/∆,œ∆ ,V /, and the symbol for a basic relative description
(o) (o)
D(o) /G(o) ,œG ,V (o) / becomes D(o) /∆(o) ,œ∆ ,V (o) /; a relative metade-
(n)
scription of order n, D(n) /G(n) ,œG ,V (n) /, n = 0, 1, 2, . . . , is symbol-
(n)
ized by D(n) /∆(n) , œ∆ , V (n) /.

Together, these symbolizations constitute the ideographic representation


→, ∆, œ , V, (D(n) , n = 0, 1, ...)] of MRC.
[↓ ∆

4.4. The Third Stage: a Scheme of a Mathematical Repre-


sentation of MRC in Terms of the Theory of Categories
The verbal formulation of MRC conveys a methodology by which the ac-
tivity of constructing knowledge, though exposed with the help of words,
nevertheless is extracted from mere language by a peculiar sort of qualita-
tive and non-mathematical formalization. The above ideographic symbol-
ization increases the degree of this extraction. But in order to increase this
liberation still more, it seems important to achieve now a mathematization.
Indeed current language inextricably incorporates hosts of surreptitious false
absolutizations, of insidious obscurities, a pullulation of sonorities and im-
plications that arouse unpredictable philosophical suspicions, refusals, pas-
sions. Furthermore, it is devoid of a clearly defined calculus. Neither a verbal
extraction via a system of definitions, postulates and axioms, nor an ideo-
graphic symbolization associated with such an extraction, cannot sufficiently
remedy to these lacunae. A transposition of the definitions and principles
which form the nucleus of MRC, in mathematical terms, would re-produce
the essence of MRC in a still more unambiguously defined form, a more syn-
thetic form, more purified of uncontrolled philosophical harmonics. It would
also open up the possibility of calculatory treatments.
On the other hand, the full content conveyed by the verbal presenta-
tion should be kept in mind: It certainly points best toward the whole wealth
of the singular conceptual being symbolized “MRC” which, like any singular
designatum, escapes any sort of language, but, if touched and grasped by
the mind in prolongation of a “direction” of thought well materialized by
associations of words from current language, acts as a guide and a fertiliser
of the process of understanding.
194 M. MUGUR-SCHÄCHTER

4.4.1. Preliminary summary

The first target of a mathematical expression is a synthetic re-expression of


the skeleton of the nucleus of MRC. So we begin by extracting this skeleton.
Imagine a consciousness functioning CF in interaction with the reality
R.

- This interaction induces inside CF epistemic aims that generate there


the conception of corresponding epistemic referentials, i.e., a priori
non restricted pairings (G, Vg ) or (G, V ).
- The epistemic aim synthesized by (G, Vg ) (or (G, V )) leads to a first
epistemic action, namely of G upon the corresponding “spot” RG from
R, that generates out of R the object-entity œG .
- Consider now the definition D7 of mutual existence of G and Vg (or
V ). If G and Vg (or V ) do not mutually exist in the sense of D7, then
the a priori pairing (G, Vg ) or (G, V ) must be a posteriori dismissed;
but if G and Vg (or V ) do not mutually exist in the sense of D7, then
the action of Vg (or V ), upon œG —to be accomplished accordingly to
the principles P8 and P10 and to the proposition π11 when spacetime
is involved—produces observable results.
- Concerning these results consider now the condition of stability from
D.14.1 (cf. also π12 and π13). If this condition does not obtain, neither
on the individual level of description nor on the probabilistic one, then
the a priori pairing (G, Vg ) or (G, V ) must be a posteriori dismissed,
even though it has resisted the first test D7 of mutual existence. But
if the condition of stability does obtain either on the individual level
of description or on the probabilistic one, then hierarchical chains of
(n)
relative descriptions D(n) /G(n) ,œG , V (n) /, n = 0, 1, 2, ... involving
(G, Vg ) or (G, V ) can be constructed accordingly to the principle of
separation P15, the concept D16 of metadescription, and to the con-
cepts D19 of intrinsic metaconceptualization; these enrich the content
of an evolving net of chains of relative descriptions.

This is the essence of the skeleton of MRC.

4.4.2. Mathematical framework in terms of the theory of cate-


gories

We seek now a mathematical representation of the skeleton of MRC. It is


crucial to begin by making use of the weakest possible mathematical struc-
ture, i.e., which introduces a minimum of formal restrictions not stemming
from MRC itself. Only in this way can it be hoped to avoid a too amputating
QUANTUM MECHANICS VS. RELATIVIZED CONCEPTUALIZATION 195

transposition of the content of the verbal presentation. Too often the formal-
izations, and in particular the mathematical ones, amputate under cover of
insuring “generality”. Later it will be useful to specify local restrictions in
order to characterize particular types of MRC-conceptualizations (logical,
probabilistic, this or that sort of theory). But the general framework has
to be maximally comprehensive. No pre-existing mathematical structure, I
think, can yield a fully satisfactory formal expression of MRC. This is so
because of the very peculiar character of the basic descriptions (D14.3.1
and D14.3.2) which introduce explicitly into the representation features re-
flecting fragments of as yet non conceptualized factuality. But the theory of
categories seems to be a good candidate for just a start. Later, once the log-
ical and the probabilistic consequences of MRC will have been exposed (cf.
Sec. 5), it will appear that probably the most specifically adequate mathe-
matical transposition of MRC will be a vectorial one. But for the moment
let us explore the expressibility in terms of the theory of categories: this
will bring into evidence certain very interesting peculiarities of MRC. To
begin with, we remind briefly of the basic definitions from the theory of
categories.
Consider the concept of “category,” as defined in Encyclopedia Uni-
versalis, Vol. 3, (France S.A., 1976), p. 1057 (my translation in which: Fl
(flèche) becomes Ar (arrow), etc.; these notations, of course, can be opti-
mized later).

A category C consists of the specification of:


(a) a class Ob(C) of objects, and a class Ar(C) of arrows;
(b) two applications s and t from Ar(C) into Ob(C) (for any pair (A, B)
of objects one denotes by Hom(A, B) the class of arrows f having the
source s(f ) = A and the target t(f ) = B; if f ∈ Hom(A, B) one writes
f : A → B, or A → B;
(c) an application that associates with any pair (g, f ) of composable ar-
rows, i.e., such that s(g) = t(f ), a composed arrow denoted g ◦ f or
gf , with source s(f ) and target t(g).
The concepts thus defined are subject to the two following axioms:
(C.1) For any object A there exists a unit arrow 1A : A → A such that
1A ◦ f = f and g ◦ 1A = g, for any arrow f with target A and any
arrow g with source A;
(C.2) If f : A → B, g : B → C and h : C → D, then (hg)f = h(gf ).
The mathematical structures (sets, groups, topological spaces, etc.)
are usually endowed with morphisms (applications, homomorphisms, con-
tinuous applications, etc.) and they determine categories (Set, Top., etc.)
196 M. MUGUR-SCHÄCHTER

whose objects are the structured sets and whose arrows are the morphisms;
the source and the target of a morphism are here, respectively, the starting
set and the arrival set of the morphism. One immediately obtains categories
that are not of the preceding type, via formal constructions like the follow-
ing ones: if C1 and C2 are two categories, the product category C1 × C2 has
as objects the pairs formed with an object from C1 and an object from C2 ,
the arrows with source (A1 , A2 ) and target (B1 , B2 ) being the pairs (f1 , f2 )
where f1 : A1 → B1 and f2 : A2 → B2 . The dual category corresponding to
a category C ∗ is obtained by “reversing“ the direction of the arrows from
C.
If C and C 0 are two categories, a functor F from C into C 0 associates
to any object A from C an object F (A) from C 0 , and to any arrow f : A → B,
an arrow F (f ) : F (A) → F (B) such that:

(F.1) for any object A from C, F (1A ) = 1F (A).


(F.2) if (g, f ) are composable in C,F (gf ) = F (g)F (f )”.

4.4.3. CMRC

Preliminaries

We shall now try to represent the skeleton of MRC, in the terms of the
theory of categories. So we shall introduce a category denoted CM RC . This
is not attempted under the constraints of the theory of models. Indeed in
consequence of the primordial role assigned in it to the consciousness func-
tioning, MRC has a strongly teleological character. Furthermore, because the
transferred descriptions root it into pure factuality, beneath language, MRC
also has a basically intensive semantic character, namely an actively created
and relative intensive character. Whereas nowadays semantics has a merely
superficial intensive character, because it starts on the level of languages and
of classical logic, so it incorporates the assumption of pre-existing and ab-
solute object-entities and predicates, and its difficulties are well-known: An
intensive semantics is not yet accomplished; even the relations to be required
between extensive and intensive semantic features are still very obscure. As
for pragmatics as a discipline incorporating teleology, it is still very incipient.
It would be at the same time hopeless and pointless to try to submit a priori
an approach like MRC, to requirements induced by other still non-stabilized
approaches that start from the current languages and from classical logic.
On the contrary, it can be hoped that a free mathematical representation of
MRC, as that one attempted below, if it succeeded, would help to build a
deep-rooted and sound extensive-intensive pragmatical semantics.
Since CM RC is attempted as a particular interpretation of the ab-
QUANTUM MECHANICS VS. RELATIVIZED CONCEPTUALIZATION 197

stract concept of a category, the semantics associated with the involved


objects and arrows will be given as much importance as the syntactical
constraints imposed by the theory of categories.

Ob(CMRC )

The objects from the class Ob(CM RC ) are called epistemic sites (in short,
sites) and are denoted S. A site is posited to designate a definite sort
of conceptual ground—just a semantic receptacle similar to an axis in a
graphic representation, or, more generally, to a multidimensional representa-
tion space—available for lodging inside it an evolving and unlimited content
to which no general structure is pre-imposed (for the representation of par-
ticular MRC-problems one can pre-impose a particularly adequate structure,
for instance an order). This content, however, is required to have a nature
consistent with the general definition of the considered semantic receptacle
(to “fit” into it, as, for instance, the red of this flower or the dark of this
cat do fit into the semantic dimension labelled by the word “colour”, but
not into that labelled by the word “form”). The most important feature of
the content of a site is that it is not required as given from the start on
(though it is permitted such): in general it is conceived of as being created
progressively and indefinitely in the same way that the “population” of a
point from an axis can indefinitely be increased by adding entities located
at that point.
The distinction itself between a stable pre-existing conceptual recep-
tacle (a genus, an axis, a multidimensional conceptual space), and a cor-
responding sort of content of which any constituent or part can always be
lodged inside this receptacle, indefinitely, at this or that definite “location”
(specific difference, point), is by no means new. Quite on the contrary, more
or less explicitly it underlies the whole classical organization of thought (lin-
guistic, logical, mathematical; it was already quite explicit for Aristotle), and
it includes also the basic notion of a referential. But neither classical logic
nor nowadays mathematics do represent in general and explicit terms the
most complete possible process of generation of the content of a pre-posited
conceptual receptacle, as it is specified inside MRC by the generator G(o)
or G(n) from, respectively, the concepts of a basic transferred description
and of subsequent intrinsic metaconceptualizations and modelizations. And
very often this content is tacitly supposed to somehow be entirely “given”
from the start on, to somehow pre-exist all done, “out there”, in a Platonic
manner. Only if ab initio this hypostatic view is systematically replaced by
a genetic one, will it be possible to mimic in the terms of the theory of
categories, the fundamental MRC-concepts of basic transferred description
and of intrinsic metaconceptualization. This is why here a specific definition
198 M. MUGUR-SCHÄCHTER

of the concept of “site” is needed.

The sites from Ob(CM RC ) are:

- SR , which represents formally the location of the evolving content of


the reality R, as defined in D2;
- SCF , which represents formally the location of the evolving content
of the consciousness-functioning CF, as defined in D1;
- Sœ , where have to be located all the formal representations of the
object-entities œG defined in D4, as these emerge;
- SD , where have to be located all the formal representations of the
relative descriptions D/G,œG ,V / (Def. D14.1) or metadescriptions
D(n) /G(n) , œ(n) , V (n) /, n = 0, 1, 2, ... (Def. D16), as these emerge.

As already stressed, the explicit distinction between a permanent site deter-


mined by a static definition, and the (in general) evolving content located on
this site, is quite essential for Ob(CM RC ). Furthermore, according to MRC it
is necessary to posit explicitly that SR ⊃ [(Ob(CM RC )] (see Def. D2), which
will induce reflexive features into the formalization [22].
In a future elaboration of particular MRC-problems, Sœ and SD will
have to be assigned structures. Sœ will have to become a mathematical space
lodging in it an evolving content of some sort of specified mathematical be-
ings (real or complex functions, kets, sequences of signs, etc.) generated one
by one and in general independently of one another and offering a convenient
representation of the considered sort of object-entities (for instance, in the
particular case of the Hilbert-Dirac formulation of quantum mechanics Sœ
becomes the Hilbert space of non-referred state vectors |ψi). SD will have
to become another kind of mathematical space, lodging in it an evolving
content of some other sort of mathematical beings, again generated one by
one and in general independently of one another and representing conve-
niently the considered type of achieved descriptions (in the case of quantum
mechanics SD consists of the space of column-matrixes that represent any
state vector in some given basis). These spaces will have to be endowed
with general structures such that the formal behaviour of the elements from
the space, if tied with physical object-entities œG , when combined with the
other elements of the mathematization, shall permit to reflect conveniently
the spacetime restrictions imposed by the principles P8 and P10, as well as
the propositions π11,π12, π13. Moreover the two structures posited on Sœ
and SD will have to be connected with one another consistently from both
a mathematical and a semantic point of view. In order to reflect formally
this or that particular class of object-entities and/or of descriptions, further
more specific structural restrictions can be added.
QUANTUM MECHANICS VS. RELATIVIZED CONCEPTUALIZATION 199

Ar(CMRC )

Consider now the class of arrows, Ar(CM RC ). The arrows from this class
will be called epistemic arrows.
Inside the theory of categories, given some category C, an arrow from
Ar(C) is currently conceived to represent an already constituted morphism
that pre-exists in a Platonian manner. This sort of semantics, however, is
not coherent with our previous definition of Ob(CM RC ) as containing sites
with evolving content. For consistency with the definitions from MRC and
with our previous definition of Ob(CM RC ), any arrow from Ar(CM RC ) will
be posited to represent a process of which the action is unlatched inside
the source-site, at a definite “content-point” which in certain cases is itself
created by that process, as its source-point; then the process develops in time
(and sometimes in spacetime) always ending by the creation at its head of
a local contribution to the evolving content of the target-site. In this sense
an CMRC-arrow is posited as a local genetic arrow.
The epistemic arrows from Ar(CM RC ) themselves are generated in-
side the consciousness functioning CF by its free choices, in consequence
of its interactions with the contents of SR and with itself. So:
Though it does not belong to Ob(CM RC ), the generic concept
Ar(CM RC ) can be best described by making use again of the con-
cept of site, namely as a site bearing an evolving content of arrows.
The set of arrows Ar(CM RC ) can be split in two sub-classes of epis-
temic arrows, a sub-class of primitive epistemic arrows P Ar(CM RC ), and
a sub-class of composed epistemic arrows CAr(CM RC ).

P Ar(CM RC ). The primitive epistemic arrows from Ar(CM RC ) are:

- Data-arrows d → denoted d, with s(d) = SR and t(d) = SCF (so belonging


to Hom(SR , SCF )), that represent the generation of data inside CF, by the
influxes from the reality R.
- Endomorphic aim-arrows, of two kinds:
* (Object-entity-generation-aim)-arrows GA → (in short GA) with
s(GA) = SCF and t(GA ) = SCF (so belonging to Hom(SCF , SCF ), that
represent the process of constitution inside CF of the aim to know
specifically about a somehow pre-figured sort of object-entity œG .
* (Qualification-aim)-arrows or, in short,view-aim-arrows, of two kinds,
Vg A → or V A →, indistinctly short-noted V A, with s(V A) = SCF
and t(V A) = SCF (so again belonging to Hom(SCF , SCF ), that repre-
sent the process of constitution inside CF of the aim to qualify (some
object-entity) via an aspect-view Vg or, respectively, a view V .
200 M. MUGUR-SCHÄCHTER

- Operational-arrows of two kinds:


* (Object-generation)-operational-arrows or, in short, generation-arrows
G → (in short G) that represent the epistemic operations of effective
generation of an object-entity. By definition s(G) = SR and t(G) = Sœ ,
so G → belongs to Hom(SR , Sœ ).
* Qualification-operational-arrows of two kinds, aspect-view arrows Vg →
or view-arrows V →, indistinctly called view-arrows (in short V ), with
s(V ) = Sœ and t(V ) = SD (so belonging to Hom(Sœ , SD )). The view-
arrows represent the elaboration of relative descriptions by operations
of qualification of an object-entity via, respectively, examination by an
aspect-view or a view. Mind that a view-arrow V→ represents globally
all the processes of examination that establish the one correspond-
ing relative description, so it has to be constructed from aspect-view-
arrows Vg → by taking into account the proposition π11.
[-] Aim-activating-arrows Aa → (in short Aa) of three kinds, that repre-
sent the passage (decided by the working consciousness functioning) from a
given epistemic aim, to the corresponding effective epistemic operation :
* (Generation-aim)-activating-arrows GAa → (in short GAa) with
s(GAa) = SCF and t(GAa) = SR , so GAa → belongs to Hom(SCF , SR );
* (View-aim)-activating-arrows V Aa → (in short V Aa) with s(V Aa)
= SCF and t(V Aa) = Sœ , so V Aa → belongs to Hom(SCF , Sœ ));
* (Descriptional-aim)-activating-arrows DAa → (in short DAa), that
just initiate globally the whole descriptional program involved in the
choice of an epistemic referential. (An arrow DA → itself, a descrip-
tional-aim-arrow, is a composed arrow and as such it will be defined
below. Nevertheless the corresponding aim-activating-arrow DAa →
is a monolithic primitive arrow with s(DAa) = SCF and t(DAa) =
SR∩D , so DAa → belongs to Hom(SCF , SR∩D ) (we have SR ⊃ SD , so
t(DAa), being in SD , is also in SR ).
- The unit-arrows required by the theory of categories for each site from
CM RC could be introduced as purely formal arrows. However it is
obvious that a fully satisfactory representation of MRC inside of the
theory of categories should endow each unit-arrow, with an adequate
semantics. This might be possible but it might involve quite non trivial
epistemological considerations. It might even lead to certain deep and
rigorous explicitations concerning the reflexive features to be assigned
to the sites from CM RC . (For SCF the role of unit-arrow could be
assigned to each one of the already defined endomorphic aim-arrows,
which involves a problem of choice). So, for the moment, we leave open
the question of a meanigful definition of the unit arrows.
QUANTUM MECHANICS VS. RELATIVIZED CONCEPTUALIZATION 201

Before continuing with the sub-class of composed epistemic arrows,


let us note the following. An epistemic referential (G, V ) as defined in D6
can be now represented formally by the corresponding pair of operational
arrows (G →, V →). In order to represent formally the a priori possibility
of any MRC-pairing (G, V ), inside CMRC any pairing (G →, V →) will be
permitted a priori. An observer-conceptor as defined in D6 can then be
represented inside CMRC by the association [CF, (G →, V →)] between the
evolving content CF of the site SCF and the representation of an epistemic
referential.

CAr(CM RC ). The composed epistemic arrows from Ar(CM RC ) are:

- Given two aim-arrows GA → and V A →, whatever they be, they are


always composable in any order, since s(GA →) = t(QA →) = s(GA →)
= t(V A →) = SCF . However the MRC-semantics requires to take into
consideration only the order GA → ◦V A →. So, denoting the result
by DA → (in short DA), we have DA with s(DA) = t(DA) = SCF . We
call it a descriptional-aim-arrow and we write

DA = DA → = GA → ◦V A →

This descriptional-aim-arrow DA → = GA → ◦V A →, like


a fragment of DNA, holds in it, still non-realized so still a-
temporal, the whole descriptional program corresponding to the
pairing (GA →, V A →), whether realizable or not.15

Given a pair of arrows d →, DA →, the composition, in this order, is al-


ways possible formally. But it is MRC-significant iff DA → corresponds
to the content of data supposed to be carried by d → (this, being a
fundamentally semantic matter, cannot be established formally). The
composition will be taken into account only when it is meaningful.
We then call it an induction arrow, we denote it ind.DA → (in short
15
The selection—among all the syntactical possibilities offered by a formalism—of ex-
clusively those that translate the semantic features to be represented, is unavoidable when
an interpretation of a formal system is built. In particular the procedure is quite current
throughout mathematical physics. (For instance, in a quantum mechanical problem of
square potentials, the general solution of the differential equation of the problem offers
exhaustively all the possible formal terms; among these, those which have no physical cor-
respondent in the data of the problem are dismissed, while the conserved expressions are
specified as required by these data (limiting or initial conditions, etc.), which cannot follow
syntactically. Another example can be found in Schrödinger’s solution of the problem of a
one dimensional harmonic oscillator where subtle and very constructed physical arguments
are introduced in order to identify restrictions that are not imposed mathematically; etc.).
202 M. MUGUR-SCHÄCHTER

ind.DA), and we write

ind.DA → = d → ◦DA →

s(ind.DA) = SR and t(ind.DA) = SCF , which represents formally an


induction of a descriptional aim from R into CF.

- Consider the representation (G →, V →) of an epistemic referential. For-


mally the two operational arrows are always composable in this order. MRC
also requires, for methodological reasons, to systematically admit the com-
posability a priori, but to exclude it a posteriori if the condition D7 of mutual
existence or the condition of individual or probabilistic stability involved by
D14, appears not to obtain. So inside CM RC we proceed as follows. First,
systematically and tentatively, we do form the composition between G →
and V →, in this order, naming it a descriptional arrow D → (in short, D).
Thus we write
D → = G → ◦V →,
with s(D) = SR and t(D) = SD (so belonging to Hom(SR , SD ). But if later
it is found that no description arises because D7 or the condition of stability
from D14 fails (which, being fundamentally a matter of semantics, cannot
follow syntactically), then we cancel a posteriori the previously formed arrow
G → ◦V → and the corresponding epistemic referential (G →, V →). Any
epistemic referential considered in what follows is supposed to have been
found to satisfy both D7 and D14. The composed arrow D → = [G → ◦V →]
formed with such a “good“ epistemic referential is the operational nucleus of
CM RC . It has to be constructed so as to yield a satisfactory formal expression
of all the conditions relevant to the considered description, as required by
D14 (so P10 and P11) as well as by (according to the case) P15, D16, D19:

In consequence of P10 and π11, D → involves an (in general) non-


commuting algebraic structure imposed upon the set of arrows V →.

- Given an epistemic referential (G →, V →), the following corresponding


composition, called a complete-description-arrow (in short, CD) is always
possible and significant:

CD → = CD = d → ◦DA → ◦DAa → ◦G → ◦V →
= indDA → ◦DAa → ◦G → ◦V →,

with s(CD) = SR and t(CD) = SD (thus belonging to Hom(SR , SD ), which


reads: Data from the reality R induce a descriptional aim into the conscious-
ness functioning, this is activated, and so first an object-entity is generated
QUANTUM MECHANICS VS. RELATIVIZED CONCEPTUALIZATION 203

out of R (which brings on the site of object-entities) and then this object-
entity is qualified, whereby a description is obtained (which brings on the
site of descriptions). The explicit “sites-trajectory“ of a complete description
CD is
SR − SCF − SCF − SCF − SR − So − SD .

The triplet SCF − SCF − SCF expresses satisfactorily the dominant role of
the consciousness functioning in a descriptional process.
- Other compositions also are permitted by the introduced definitions (for
instance GAa → ◦G →, V Aa → ◦V →, etc.). But it seems not necessary
to examine them exhaustively. Notice that the MRC-definition D2 of real-
ity requires to extend now the previous assumption SR ⊃ [(Ob(CM RC )] by
positing explicitly SR ⊃ [(Ob(CM RC )+Ar(CM RC )].
The axioms C1 and C2 . They seem to raise no problems.

Representation of the evolving contents of the CM RC -sites.


The theory of categories does not specify a general modality for ex-
pressing individualizations inside an object from Ob(C), as being the source
or the target of an arrow tied with that object. While MRC involves such
individualizations quite essentially. So we construct the necessary individu-
alizations as follows.
We consider only the operational arrows G → and Vg → that form
the hard core of CM RC . This will suffice.
Each arrow G → can be labelled by a pair of indexes (RG , œG ) defining
respectively its local start inside SR (by the “spot“ RG where G has to
be applied (D4)) and the element œG from the evolving set {œG } that
constitutes the new contribution to the content of Sœ by the creation of
which the considered G → arrow ends. So for each definite arrow G → we
shall write (RG , œG ) →, which distinguishes it from any other arrow G →.
Thereby the set {(RG, œG ) →} associated to the generation arrows G →,
itself also an evolving set, is now connected with the evolving inner contents
of the two sites SR and Sœ represented, respectively, by the evolving sets
{RG } and {œG }. This connection can be then organized more by putting
mutually compatible structures on the sets {RG }, {œG } and {(RG , œG ) →}
(physical operations of object-entity generation are subject to the frame-
principle P8, which requires a convenient extension of the principle P10 of
mutual exclusion, to operations of object-entity generation also).
Mutuatis mutandis, one can connect in a similar way each definite
processual arrow Vg →, with a “pair” of indexes (œG , {gk}), k = 1, 2, ...,
by re-writing (œG , {gk}) →, k = 1, 2, . . . ,where k takes on a unique value
if the attempted descriptional process reveals an individual stability, or a
whole set of different values if it reveals a probabilistic stability ((D5.1),
204 M. MUGUR-SCHÄCHTER

π12, π13, D14). In (œG , {gk}) the index œG defines the element from the
discrete evolving content of the source-site Sœ where (œG , {gk}) → begins,
and {gk}, k = 1, 2, ... defines the element from the discrete evolving content
of SD by the creation of which (œG , {gk}) → ends. So the (evolving) set
{(œG , {gk}) →} of aspect-view arrows is connected with the evolving content
of the sites Sœ and SD , expressed respectively by the sets {œG } and {gk}
(where {gk}, k = 1, 2, ..., g fixed, amounts to the description of œG via Vg ,
which is an element from {D}). The connection between the evolving sets
{œG }, {(œG , {gk}) →} and {D} can be then organized more, by putting on
these sets mutually compatible structures obeying all the MRC-requirements
and furthermore conveniently reflecting the particular considered class of
descriptional processes (the nature presupposed for the object-entities and
the aspect-view-examinations).
The procedure can be extended to the class of arrows V →: in con-
sequence of D5.2 each definite V → arrow is a set of arrows {(œG , {gk}) →,
k = 1, 2, . . . , }, g = 1, 2, . . . m, m finite.
Then a relative description D/G,œG ,V / from MRC becomes in CM RC
a complete-description-arrow

[CD → = CD = d → ◦DA → ◦DAa → ◦G → ◦V →],

where G → ◦V → is indexed:

(RG, œG ) → ◦(œG , {gk}) →, k = 1, 2, ...}, g = 1, 2, . . . , m, m finite.

CM RC versus quantum mechanics.

We consider the Hilbert-Dirac formalism of quantum mechanics. The


Hilbert-space H of the state-ket-vectors |ψ > of the studied microsystem
corresponds to the CM RC -site Sœ where are lodged mathematical represen-
tations of the considered class of object-entities. The set {|ψ >} of state-
ket-vectors |ψ > from H corresponds to the evolving set {œG } from Sœ .
The vector-space structure assigned in quantum mechanics to {|ψ >} is a
particular feature entailed by the principle of superposition posited for quan-
tum states, a principle justified by the wavelike features manifested by what
is called quantum states. So in general such a structure has no semantical
counterpart, so it will have to be dropped. Furthermore

The CM RC generation arrows (RG , œG ) → have no general corre-


spondent in the quantum mechanical formalism: they are represented
only in the particular case of microstate-generation by a measurement
process.
QUANTUM MECHANICS VS. RELATIVIZED CONCEPTUALIZATION 205

This is a striking lacuna (which is suppressed in meta[quantum me-


chanics]) (see note 1).
The quantum mechanical (in general) non-commuting linear differ-
ential “dynamical” operators defined on H correspond to the CM RC -aspect-
view arrows (œG , gk) →, k = 1, 2, . . . .
The quantum mechanical representation of a state-ket |ψ> with re-
spect to the basis of eigenvectors introduced by a given quantum mechanical
operator A, namely as a column-matrix of which the elements are calculated
with the help of |ψ > and the considered eigenvectors, corresponds to a ba-
(o)
sic transferred description D(o) /G(o) ,œ(o) ,Vg / from SD created for a basic
object-entity œ(o) by a basic aspect-view-arrow (œG , {gk}) →, k = 1, 2, . . .
(which can be re-written (œ(o) , gk (o) ) →, k = 1, 2, ....).
The set of all the column-matrix representations of a given state-ket
|ψ> with respect to all the bases of eigenvectors introduced by all the quan-
tum mechanical dynamical operators, corresponds in CM RC to a complete-
description-arrow

CD → = CD = d → ◦DA → ◦DAa → ◦G → ◦V →

(with G → ◦V → indexed: (RG , œG ) → ◦(œG , {gk}) →, k = 1, 2, ...}, g =


1, 2, ...m, m finite).

So it will be possible to attempt a systematic transposition of the


Hilbert-Dirac formulation of quantum mechanics, in terms of the
theory of categories, via MRC with its central concept of basic trans-
ferred description.

It is of course obvious from the start that the explicit CM RC -


representations of reality and of the consciousness-functionings have no cor-
respondent in quantum mechanics where not even the actions of object-
entity generation are represented mathematically, nor are they at least con-
ceptually and verbally clearly distinguished from the qualifying actions via
measurements. By comparison with CM RC quantum mechanics appears as
flawed by very flattening lacunae.
Nevertheless, once the main relations CM RC -(quantum mechanics)
have been established, the quantum mechanical formalism becomes a pre-
cious guide for a subsequent development of CM RC (any non-necessary re-
striction suggested by the—particular—case of quantum mechanics having
to be carefully avoided). One first important step in the mentioned direction
will be the identification of the individualized MRC-meaning of Dirac’s dual
space of linear functionals defined on the Hilbert space of state-ket-vectors,
and of the various sorts of scalar products from the Hilbert-Dirac formula-
206 M. MUGUR-SCHÄCHTER

tion of quantum mechanics (see [13]). Then the CM RC -transposition of these


will have to be conveniently achieved.

4.4.4. Concluding comment on CM RC

The outline indicated above needs development. For instance, the condition
SR ⊃ [(Ob(CM RC )+Ar(CM RC )] imposed by MRC entails reflexive characters
that might raise difficult syntactical problems connected with the definition
of the categorial concept of a sub-object. The postulate, the principles and
the propositions from MRC must systematically acquire inside CM RC math-
ematical expressions, and the MRC-propositions should furthermore acquire
mathematical proofs. Etc.

5. ILLUSTRATIONS OF THE FUNCTIONING OF MRC

In this section we illustrate by examples the functioning of MRC, thereby


also developing the method. We shall first consider logic and then probabil-
ities.

5.1. Classical Logic [23] versus the MRC-logic of Relative


Classes of Cognitive Actions
Because logic is so particularly important when a method of conceptualiza-
tion is proposed, we shall, by a brief sequence of remarks, try to convey a
notion concerning the relations, and the gap, between MRC and classical
logic. We shall then very briefly indicate along what lines an MRC-logic
can be constructed and what novelties it introduces. It will appear that the
MRC-logic achieves an explicit connection between physical factuality and
formal structure, and that it disconnects the question of the consistency of a
formal system, from the question of decidability (completeness) considered
in Gödel’s basic theorem, on which it yields a different perspective.

5.1.1. Critical remarks on Frege’s basic definitions

Insufficiencies of the concept of Frege-class of a predicate

The logic of classes and predicates has first been developed by Frege. The
starting remark is that a predicate “determines“ a class of objects, namely
those that partake of the meaning (sense, comprehension) of the considered
predicate and hence constitute its extension. In order to identify these ob-
jects, first (a) it is remarked that a predicate, by itself, is neither true nor
false, but that (b) its assertion concerning a given object-entity can be true
or false if the predicate is “pertinent” concerning this object-entity. Then
QUANTUM MECHANICS VS. RELATIVIZED CONCEPTUALIZATION 207

(c) for each predicate P a propositional function fP (x) is introduced where


fP represents the predicate and x is an object-variable:

“The expressions which.....include letters ‘x’,‘y’,‘z’, and are such that


they become true or false propositions as soon as the objects desig-
nated by these letters are specified, are called propositional functions
(J-B. Grize [23], p. 150).”

And (d) it is posited by definition that any value of the object-variable x


for which fP (x) is true, belongs to the class determined by P . In short:

The class of P is the set of values of the object-variable x for which


fP (x) is true.

From the standpoint of MRC these very first steps call forth already
the following remarks:

* In the first place, we are in presence of a qualification16 of an object-


entity—“the class of P ”—of which the generator G is of a particular and so
a restrictive type, namely a “generator G(V ) of a view V ” (cf. final general
comments on D14 and 5.1.2): V is supposed to act first in the role of a gen-
erator G(V ) that selects as object-entity the whole field of perceptibility of
V (“any value of an object-variable...”), and then it furthermore acts in the
role of a view, by qualifying isolately the “values of the object-variable” from
this field of perceptibility, but qualifying them from inside the metapredicate
“P is true” (cf. the sequel). This very particular sort of generator of object-
entity, G(V ), produces either conceptual object-entities—i.e., already pre-
viously achieved descriptions—or basic object-entities that transfer directly
on the sensitive biological apparatuses of the human beings, marks called
“impressions”. It has been already remarked that this last sort of cognitive
situation produces basic transferred descriptions that are spontaneously and
implicitly metaconceptualized during the very first period of a man’s life,
and are reduced to intrinsic models (D19.3) that seem to pre-exist indepen-
dently of observation, “out there”, available for examinations, in particular
examinations of [truths of P ’s].

Both sorts of descriptions mentioned above, perpetuate a full igno-


rance of the rooting of conceptualization, in physical factuality.
16
I say “qualification”, not “description”, because no condition of stability of the qual-
ificational result with respect to repetitions of the process of qualification, is required
here, as it is in the set of all the definitions D14 (with the unique exception of relative
testimonies D14.2.2) (cf. the comments on the generalization of D14.1).
208 M. MUGUR-SCHÄCHTER

* In the second place, the qualification “x is P ” and the metaquali-


fication of empirical truth of this first qualification, are combined in a sort
of coalescence where fundamental MRC-conditions get lost. Indeed from the
point of view of MRC the qualification “x is P ” is just a piece of meaning,
no matter whether true or false, and possibly not even that, if a posteriori
it appears that no view of empirical truth can be constructed which exist in
the sense of D7 with respect to the assertion “x is P ”. Whereas in Frege’s
approach such reservations are totally absent. Moreover the qualification “x
is P ” is first introduced in a quasi subliminal way, and then it comes into
stable being together with, and indistinctly from the metaqualification “it
is true (or false) that (x is P )”. This conveys the illusory assumption that
a truth-qualification is always possible for any qualification “x is P ”, what-
ever its semantic content. Which of course is not the case, as Tarski claimed
much later (“the snow is white” is true iff the snow is white).

* In the third place, the involved predicate P , considered separately,


is neither endowed with some structure, nor is it subjected to any condi-
tions of effectivity of the examination which P is supposed to perform on
x: a sort of ghost-predicate (compare with an aspect-view D5.1 or a view
D5.2). Furthermore, as just mentioned, the so feebly formed significance of
what is called a predicate P is immediately dissolved in the metapredicate
of [truth of P ]. While for the metapredicate of [truth of P ], again, no struc-
ture whatever is specified, nor some condition of relative existence and of
effectivity.

In sum, on the one hand, a predicate P and its truth qualifications


are assigned the fundamental logical role of, together, producing always,
automatically, a proposition, i.e., the tentative assertion of a description,
that can then be found, via some definite procedure, to be true or false. But
on the other hand:

The classical predicates “P ” are reduced to no more than shadows of


undefined intensive extracts from factuality, just verbal labels which,
while they are hypostatized, are also smuggled away by an immediate
translation in terms of a purely extensive domain of correlates “x”
inside the realm of object-entities on which they act, this correlation
being subjected to another undefined meta-intension called truth. A
vague but dense knot.

* Consider now the “values of x” in general—not only those selected


in “classes of P ”—and notice that these are the equivalents of MRC-object-
entities œG . Now, no genesis whatever is specified for the “values of x”. They
QUANTUM MECHANICS VS. RELATIVIZED CONCEPTUALIZATION 209

are simply posited to always be “out there”, passively waiting to fall inside
the field of perceptibility of the predicates P .

Classical logic implies in its foundations a hypothesis of


universal actuality.The Boolean algebra of classes and predicates
is constructed for the already actualized.

(This, by isomorphy, holds also concerning the nowadays set-calculus on


which classical mathematics are founded). Fundamentally, the modal dimen-
sion of existence (not to be identified with the “logical” modalities of neces-
sity or possibility) along which potential existence is transformed into actu-
alities by processes of actualization, remains exterior to the classical calculus
of classes and predicates. When needed, this dimension has to be superposed
by a posteriori manipulations. This is not disturbing in the usual language
where everything is plethoric, contextual, minimally structured, which for
the specific aims of current language is optimal. But in a fundamental for-
malized representation of thought operations, like logic, the absence of the
modal dimension of existence is an imprisoning poverty comparable to what
the absence of techniques for the representation of perspective must have
been in painting. Only addiction to the traditional methods can hinder to
perceive to what a degree such a lacuna is amputating, and that, in partic-
ular, it is an obstacle in the way of a basic and explicit connection of logic,
to conceptual geneses, to aims (finality), to praxis.

* The fact that no genesis is specified for the “values of x” (the


MRC-object-entities œG ) has also another consequence, a radical one:

The generators G of the object-entities œG themselves (not the gen-


erators P of the [classes of some P ]) are simply not considered.

This absence of an explicitly defined object-entity generator G, so, a


fortiori, the absence of a generator G required to be in general independent
of any qualification and permitted to be physical-operational, restricts a
priori and arbitrarily the domain of object-entities to which the classical
logic can be applied:

In classical logic all the basic physical object-entities that have to be


first radically generated by deliberate physical operations of object-
entity-generation, independently of any subsequent qualification, and
then might have to be transformed in order to draw from them ob-
servable manifestations, are simply eliminated a priori from consid-
eration. Indeed “predicates”, i.e., linguistic-conceptual qualificators,
cannot “determine classes” among basic object-entities in the sense of
210 M. MUGUR-SCHÄCHTER

MRC. They cannot act upon such only factually singularized object-
entities, because they are not homogeneous in nature with these.

Since the cognitive situation tied with basic physical object-entities, how-
ever, is endowed with a certain universality of principle (3.3), we are in
presence here of a huge arbitrary amputation. Namely the massive ampu-
tation of the whole stratum of conceptualization where the structure of its
rooting in physical factuality is specified. In such conditions one can, in
particular, well understand why, for classical rationality, quantum mechan-
ics seems unintelligible. Indeed one of the fundamental features of quantum
mechanics is precisely the liberation (in general) of the operation of gener-
ation of object-entity from any view. And it is by this liberation that MRC
transpierces the armoured platform of language and succeeds to build a rep-
resentation of the processes of conceptualization that is rooted in physical
factuality. But, and this comes as a surprise, not exclusively basic physical
object-entities are eliminated because the object-entity generators are not
explicitly considered. All the conceptual entities which are first constructed
independently of any qualification and are only afterwards qualified—like
many mathematical systems and formal systems of logic itself !—are equally
eliminated from clear consideration as created, constructed object-entities.
This leads to false problems, and to enormous unnecessary efforts to solve
them (5.1.2). An amputation of such an extent, and which concerns logic
itself, is not acceptable in a fundamental discipline like logic.

Let us now take a second step. By definition:

“Two classes of object-entities α and β are equal iff all the elements
of α are elements of β and vice versa (iff α and β hold the same elements).
Two propositional functions fα et fβ that determine two classes α
and β are equivalent if the classes α and β are equal (cf. op. cit., in contin-
uation).”17

This prompts a new critical remark:

* How can one know, for instance, whether yes or not for any value of an
x for which it is true that it is red, it is equally true that it is spherical?
It is implicitly supposed that the answer to such questions can always be
given. But this supposition is founded upon the same restrictive hypothesis
identified above that any value of any object-variable x (an object-entity
17
Hervé Barreau remarked that precisely these definitions have already been the object
of basic criticisms opposed to Frege’s logic. This might somehow be related to the remarks
that follow in the main text. However here Frege’s approach is examined exclusively by
confrontation with MRC, and on a level of principle where technical features do not appear.
QUANTUM MECHANICS VS. RELATIVIZED CONCEPTUALIZATION 211

œG in MRC terms) pre-exists out there, already accomplished, ready to


be pointed toward with one’s finger, certainly available for examination via
the metapredicate [truth of P ’s], equally always available. But this time
it is furthermore implied that a P -examination of a value of an x never
changes the considered value of that x: if it did, this value of the considered
x, after having been examined by P ≡ α, would in general cease to stay
available for an examination also by P ≡ β. Thereby, again but otherwise,
are eliminated a priori all the basic transferred descriptions that are so
deeply rooted in physical factuality that they have to be radically changed in
order to draw from them observable manifestations. Now, in the constructive
outline from Sec. 5.1.2 it will be shown that the basic descriptions, precisely
because in general they have to allow for changes of the involved basic object-
entity during its processes of qualification, entail certain consequences on
logical form, consequences involving strict physical singularity. But when the
rooting of logic in the as yet unknown physical factuality is obtruded, these
consequences remain hidden with it, which restricts a priori our perception
of logical form, to exclusively its plural, statistical aspects.
* This radical occultation, in classical logic, of the features tied with
strict physical individuality, is what permitted to claim that logic is just
formal structure; more, to require logic, for the sake of “maximal generality”,
to be a “pure” syntax, freed of any intension, cut from any semantic matter.

But in fact this severance is illusory. It has been possible to imagine


it to be realizable precisely because the way in which unspeakable factuality
loaded with semantic potentialities is drawn into descriptions at each local
relative zero-point of a descriptional chain, remained so completely ignored.
As soon as one becomes aware that any local zero-point contributing to
the foundation of descriptional chains, consists of a (more or less canonical)
transferred description, the illusion of the possibility of a complete elimina-
tion from a syntax, of any semantic content, is dissolved. It becomes clear
that any syntax stems from numerous bulks of physical factuality, which is
the prime matter for phenomenal appearances. It is out of these bulks that
are drawn the observable manifestations of which the phenomenal appear-
ances consist, while the whole conceptualization is founded on phenomenal
appearances. Through these phenomenal appearances, semantic matter goes
over into language-and-conceptualization, by primary codings, and then it ir-
repressibly diffuses up into all the levels of abstraction and complexification.
Language is a circulatory-system for factual, semantizable prime matter. It
emerged and got form in order to carry from mind to mind information
about factuality, about semantics. If this were not so the societies of men
would not have lasted. They would not even have started being. Our minds
212 M. MUGUR-SCHÄCHTER

work with intensions. These, adduced inside language by the phenomenal


effects of the interactions between pure factuality and mind, have then os-
motically impregnated with semantic contents all the levels of abstraction.
So, unavoidably, they have infused into logic also, where they generate its
natural forms,18 those which, more or less implicitly, command in real cir-
cumstances our choices, methods, and actions. If on the other hand in the
theoretical formalized logic any connection between syntactical form and
semantic content is first refused, this instils there by reaction lacunae and
awkward features as well as difficult fictitious problems, like for instance
those of the a posteriori connectivity of modern formal logic with modern
semantics. (Try to design in abstracto the human circulatory system, strictly
without using as a guide the condition that blood has to circulate in it in
such a way as to nourish every tiny volume of living tissue: what is the
chance to end up with the natural scheme?).

Criticism of the classical logical void ∅: the semantic relativities


of ∅

A trivial but striking example of the way in which ignorance of semantic


aspects induces syntactic insufficiencies, is that of equality of all the void
classes. In consequence of the extensive (set-theoretic) definition of the
class determined by a predicate, all the void classes are equal because they
all contain “the same element”, namely the null element ∅. So, if it is true
that no immortal man exists, and it is equally true that no symphony lasting
less than one minute does exist, then the class of immortal men and that
of symphonies shorter than one minute, are equal. This argument induces
a feeling of artifice, of twisting of what one would be prepared to accept
as “meaningful”. One feels a gliding. The trajectory of this gliding can be
retraced:
When one wants to determine quantity, extension, number of el-
ements, starting from only the quality—the predicate—that qualifies, a
ground for ambiguity is surreptitiously inserted. So long as a class in the
sense of Frege is not void, the quality specific of this class—the one expressed
by the predicate P that determines the class—is present, it is held by each
element of the class. But at the limit where the class becomes void, the
specific quality P that characterizes the class is discontinuously transmuted
into pure qualitatively indistinct quantity, into a purely numerical zero. This
transmutation has been instilled as follows. The mathematicians, when they
defined the number zero, in fact have extrapolated into nothingness a cer-
tain quality, namely the degree of “numericity” N of any (finite) number,

18
Cf. [21].
QUANTUM MECHANICS VS. RELATIVIZED CONCEPTUALIZATION 213

so the predicate P 0 ≡ N of which mathematics studies the manifestations


via the object-entities called “numbers”, of which a quite general property
is to be able to “measure”, to quantify. It is the prolongation into nothing-
ness, of numericity, of this particular predicate P 0 ≡ N , which, by definition,
has been called “the number zero”. Whereas the logicians, while they make
use of numbers in order to measure by the help of P 0 ≡ N the quantity of
supports of a given quality P 6= P 0 —this time any quality P whatever, any
predicate – did not take care to prolong into nothingness also this quality
P , in order to dispose of a veil of quality P , specifically, to be co-extended,
together with the mathematician’s zero-of-numericity, over the void encoun-
tered at the limit where the quantity of carriers of this quality P comes to
an end. So at that limit they are left with only a zero-of-numericity, un-
covered, stripped of quality P . While the other numbers of carriers, 5, 100,
etc., were all tied with also the quality P characteristic of the considered
class: at this limiting point, the conservation of the way of representing a
class breaks down, a solution of continuity inside the way of representing
a class has been surreptitiously introduced. This is a heavy methodologi-
cal error, comparable, for instance, to a dimensional inhomogeneity inside
an equation. The non homogeneity of conceptual treatment inside a closed
conceptual system is always the source of very slippery problems. Any two
void classes are considered to be “equal” on the basis of a purely extensive
estimation of the null content of a concept that has been first characterized
in an exclusively intensive way, even if this characterization possessed also
an extensive counterpart: a predicate P is only quality, and, by definition,
it is P alone that determines the corresponding class fP , not also the quan-
tity of carriers of the quality P . It is then inconsistent, if one distinguishes
clearly between quality and [quantity of supports of this quality] (in MRC
terms between views V and object-entities œG that exist in the sense of
D7 with respect to this view), to permit the defining quality to disappear
“because” all its supports disappeared, while the class itself, defined by
the quality, is still maintained. The predicate P that defines the class fP
should subsist with the class, in spite of the vanishing-support-of-quality-P,
i.e., when the set of numbers that label the supports reduces to the number
0. It is inconstant to end up in such a materialist idolatrous manner when
one has begun by adoring an abstract God. One should act like the mathe-
maticians, or like Lewis Carrol who leaves us with smile-of-cat-without-cat
when the smiling- cat vanishes completely.

The logical void ∅, which is an element of the “purely” syntactical sys-


tem called the classical logic of classes and predicates—is subjected to
semantic relativities that require a specific syntactical expression:
the asserted possibility of a radical separation between syntaxis and
214 M. MUGUR-SCHÄCHTER

semantics is obviously contradicted in the case of the logical void.

Ferdinand Gonseth said that “logic is the physics of any object.”


But any given object has some semantic content, and the types of semantic
content have to be mutually distinguished in a thoroughly worked out formal
representation of safe derivational vehiculations of our knowledge concerning
empirical truth-valuations involving “any” object.

Global critical conclusion

The classical logic of classes and predicates, which founds the whole modern
classical logic, floats above language, inside the stratum of the already pre-
verbalized-conceptualized. The rooting of the processes of conceptualization,
in physical factuality, the creative cognitive actions which produce object-
entities and qualificators of these, the modal dimension of existence where
potentiality, actualization and actuality are located, remain hidden to it. By
occultation of the genetic stages from the processes of conceptualization and
by substitution to these of false hypostatizing absolutizations, it introduces
arbitrarily restricted conceptual platforms that cannot withstand artificial
and inadequate formal representations.
Only when all the involved descriptional geneses, with the descrip-
tional relativities entailed by them, are explicitly taken into account, is it
possible to dominate from a formal point of view any descriptional situation,
whatever its complexity. This can be better understood per a contrario and
on examples.
For instance, inside MRC where any descriptional relativity is taken
into account explicitly as soon as it comes into play, the treatment of the
logical void is preorganized in consequence of the way in which the very first
levels of general conceptualization are structured. As soon as one considers
an (independently defined) object-entity œG and a view V (D4 and D5),
the test of their mutual existence in the sense of D7 is methodologically
required, before trying to perform the corresponding relative description.
If this test is negative one finds oneself precisely in the case that can be
designated as “the void class determined by V inside the set of object-
entities œG ”, which means “absence of object-entities œG admitting of the
qualification V ”, i.e., absence of the possibility of a meaning generated by
the pair (G, V ). So a conceptual void, doubly relativized to the semantic
features involved by the considered pair (G, V ), comes into being ab initio.
Later, once the possibility of meaning has been insured by mutual existence
in the sense of D7 and then a first descriptional level has been insured by the
existence of some stability of the qualifications in the sense of D14.1, comes
furthermore into consideration, in its turn, the question of empirical truth:
QUANTUM MECHANICS VS. RELATIVIZED CONCEPTUALIZATION 215

given an already achieved description in the sense of one of the definitions


D14, is this description a proposition, i.e., does it exist in the sense of D7 with
respect to some view of empirical truth that can be effectively exhibited? The
still higher and more particular level of “logical” characterizations concerns
object-entities consisting of systems of propositions. A proposition from a
formal system of propositions S, can be described by the logical views of
provability inside S and of decidability inside S, while the system itself
considered as a whole can be examined by the logical views of completeness
and of formal consistency. All the mentioned sorts of logical description are
related with the previously developed relativized conceptual-semantic voids
(mutual exclusions, absence of descriptional stability). Indeed these entail
the definibility of syntactical, calculational relativized voids (see Sec. 5.1.2)
and thus they go over into the form of the logical descriptions. So in this
specific case it is clear that, and how, inside MRC the semantic contents
determine progressively aspects of logical form. And these, the calculational
relativized logical voids, preserve from a whole category of false problems.
Indeed the absolutization of the logical void is one of the most prolific sources
of illusory problems. (Even in modern quantum logic there subsists much
confusion concerning complementations tied to the logical void [24]; cf. also
[13] as well as 5.1.2).
When instead of a system of propositions, a formal system in the
most abstract sense is considered, either any connection between semantics
and syntax has deliberately been suppressed by the process of conceptualiza-
tion (which is difficult) and in this case one obtains just a Wittgensteinian
“game” that resists any non distorting and useful interpretation in terms of
some domain of natural facts, or some connections between semantics and
the constructed formal system have been deliberately preserved, and then
precisely these insure possibilities of useful interpretations of this system.
The corpus of relativizations required by MRC does not only insure
a controlled penetration of semantics into the logical descriptions, it also ex-
erts another crucial sort of control which classical logic cannot exert system-
atically because of the artificial separation between semantics and syntax.
Namely, it insures automatically all the types of descriptional “homogene-
ity” amounting to the conservation of the method of representation inside a
closed descriptional universe, i.e., throughout the work accomplished with a
given epistemic referential. While, on the other hand, the principle of sepa-
ration P15 regulates the passages from one set of homogeneous descriptional
contents, to another one whatever their type.
This is important. Indeed the creation of sense, in all its stages, is
ruled by the implicit imposition of methodological principles of homogene-
ity: physical operations can directly change only physical entities, concepts
216 M. MUGUR-SCHÄCHTER

can directly change only concepts and can be localized only inside nets of
concepts; in an equation the semantic dimensions from the first member
must be the same as the semantic dimensions from the second member;
statistical-probabilistic qualifications do not exist in the sense of D7 with
respect to individual events, nor with respect to only statistical distributions
of events, they exist only with respect to statistical-probabilistic distribu-
tions of events; and vice versa, individual qualifications do not exist in the
sense of D7 with respect to statistical-probabilistic distributions, they are
blind with respect to these; etc. When no matter which one among these
various sorts of implicit principles of homogeneity is violated, paradoxes
or false problems emerge. Inside MRC this is always expressed as a conse-
quence of a violation of the principle of separation P15, i.e., of a non-explicit
modification of the epistemic referential which is made use of.
The false absolutizations that flaw the classical logic of classes and
predicates have prolongations in many domains of modern science, in partic-
ular in the theory of sets. Indeed the elements of a set are always supposed to
somehow pre-exist already realized, and this, just like in the definition of the
equality of two classes and of the equivalence of two propositional functions,
entails arbitrary a priori restrictions. But the most noteworthy consequence
might consist of the fact that classical logic, because of its lack of explicit
connection with strictly singular physical factuality, remains unaware of the
spacetime specificities of the descriptions of physical object-entities. This has
favoured a surreptitious gliding

conceptualization→natural logic→ formal logic→calculus→computation,

whereby often, in computational simulations of physical processes, under-


standing disappears entirely into mere doing.
Globally, the apparently so clean-cut and rigorous classical logic,
when scrutinized, reveals non-intelligibility tied with superficialities and ar-
bitrary posits.

5.1.2. Outline of an MRC relativized genetic logic

The preceding critical considerations entail by contrast a constructive ap-


proach of which what follows conveys only an extremely synthetic notion.
The aim is, inside MRC, to explicate the consequences upon logical descrip-
tions, of the relativization to the cognitive actions from which these logical
descriptions stem, so also to the semantic contents introduced by these cog-
nitive actions. The main step is the introduction of the concept of genetic
class.
QUANTUM MECHANICS VS. RELATIVIZED CONCEPTUALIZATION 217

Double extremity genetic classes

Let us recall that inside MRC what is called “object-entity” is just a de-
scriptional role (see final comment on the general concept D14 of a relative
description). No entity never pre-exists as an object-entity. It always has to
be introduced in the role of object-entity by the explicit action of a definite
operation G which either radically creates—physically or conceptually – an
actor for this role, or only recruits some pre-existing entity for acting in this
role. This, in general, is done independently of any pre-established quali-
fication of, specifically, the object-entity introduced as such by the chosen
operation G, so also independently of any “predicate P ” (cf. comment on
D14.3.1). Only after having been thus put, via G, in the role of object-entity,
becomes the involved entity available for the action on it of a view (D5.1,
D5.2)—any one—which, in its own turn, is chosen for acting in the role of
a view. The necessity of an apparently so redundant and intricate way of
saying can be best understood when the chosen epistemic referential has the
particular degenerate form (G(V ), V ) where G(V ) is the “generator of the
view V ” (cf. the general final comment on D14; V denotes here indistinctly
an aspect-view or a view while if specifically an aspect-view is meant, we
write Vg ) which is precisely the form presupposed implicitly by the whole
classical logic. Indeed in this case the view symbolized by V , though from
the beginning on it is structured accordingly to the definitions D5.1, D5.2
of qualificators, nevertheless acts first in the role denoted G(V ) of generator
of object-entity. Namely it acts first either by selecting as object-entity its
own field of perceptibility, or by radically creating this field, like for instance
in the case of the generation of a microstate by a given quantum mechani-
cal measurement process. And afterward, on the product of this first action
accomplished by itself but in the role G(V ) of generator of object-entity, V
can furthermore act also in the role of a view or an aspect-view, for which
its initial definition has been specifically intended. (Let us also recall that
what is structured as an aspect-view or a view in the sense of D5.1 or D5.2,
respectively, can be selected for the role of object-entity, by a convenient gen-
erator (here a conceptual selector) (cf. the final general comment on D14).)
The existence of situations like those mentioned above requires indeed ways
of speaking that distinguish clearly between the general descriptional roles,
and the specific actors to which the roles are assigned. This distinction is
quite essential because according to MRC, in order to describe, both the
role G and the role V have always to be acted, even if in a reduced or a
degenerate way and which reflects also the characteristics of the particular
actor put to hold the role. So inside MRC it would be neither necessary
nor sufficient to consider, as it might seem natural at a first sight, that the
218 M. MUGUR-SCHÄCHTER

equivalent of a “predicate P ” is just an aspect-view Vg . In order to achieve


qualifications, MRC requires to make systematically use, instead of just a
“predicate P ”, of some definite succession [G.Vg ] or [G.V ] of an actor put
in the role G followed by an actor put in the role V .
So far the relative description D/G,œG ,V / produced by an epistemic
referential (G, V ), once obtained accordingly to one of the definitions D14,
has been considered separately from its genesis. By the following definition
DL.1 (L: logical) we shall now introduce a synthetic concept that takes
systematically into explicit account, together with a given description, also
the whole genesis involved by it.

DL.1. Double-extremity genetic classes. Consider an epistemic


referential (G, V ) where V is a view containing in general several aspect-
views and which exists in the sense of D7 with respect to the generator G
of object-entity.

DL.1.1. Double-extremity genetic class involving a physical


object-entity. Suppose that (G, V ) introduces a physical object-entity and
that it does produce a relative description D/G,œG ,V / of it in the sense of
the definition D14.1, individual or probabilistic. Then the repetitions of the
succession [G.V ] of pairs of cognitive actions, constitute [the class of all the
operational processes of gk-valuations involved by D/G,œG ,V /] (in this con-
text the term “operational” is intended to stress that no model whatever is
asserted). The class specified above will be called a double extremity genetic
class involving a physical object-entity, in short a physical genetic class, and
will be labelled Cph [G.V ].
If V consists of only one aspect-view Vg an aspect-description
D/G,œG ,Vg / is obtained, and the succession [G.Vg ] produces [the class of
all the operational processes of gk-valuations involved by D/G,œG ,Vg /]. We
name this a one aspect double-extremity genetic class involving a physical
object-entity, in short a one aspect physical genetic class, and we label it
Cph [G.Vg ].
When a basic referential (G(o) ,V (o) ) is considered, a basic transferred
description D(o) /G(o) ,œ(o) , G,V (o) / is obtained, and the corresponding ge-
netic class will be called a basic genetic class; such a class will be denoted
C[G(o) .V (o) ]. This is [the class of all the operational basic processes of gk-
(o)
valuations involved by D(o) /G(o) ,œG ,V (o) /].

DL.1.2. Genetic class involving a non physical public object-


entity. Suppose that (G, V ) introduces a non physical public object-entity
and that it does produce a relative description NPP.D/G,œG ,V / of it in the
sense of D14.2.1. Then the repetitions of the succession [G.V ] constitute [the
QUANTUM MECHANICS VS. RELATIVIZED CONCEPTUALIZATION 219

class of all the processes of gk-valuations involved by NPP.D/G,œG ,V /]. The


class specified above will be called a genetic class involving a non physical
public object-entity, in short a non physical public class, and will be labelled
CNPP[G.V ]. In particular V can consist of only one aspect-view Vg and then
we have a class CNPP[G.Vg ].

DL.1.3. Testimonial double-extremity genetic class. Suppose


that (G, V ) does not insure the possibility to realize arbitrarily many rep-
etitions of the successions [G.Vg ], for all the aspect-views Vg from V . So it
produces an only testimonial description θ/G,œG ,V / in the sense of D14.2.2.
Nevertheless according to MRC in this case also a certain set of known or
unknown implicit qualifying processes [G.Vg ] have necessarily been involved:
if not, there would be no qualification at all (cf. general comments on the
definitions D14). These will be said to constitute the double-extremity ge-
netic class of the testimony θ. Such a class will be indicated by the notation
Cθ (G, V ) where only the involved epistemic referential is specified.

Comment. The general concept of a genetic class is posited here as


the MRC-equivalent of the Frege-class of a predicate P .
The distinction between a relative description D/G,œG ,V / (or a rel-
ative testimony)—i.e., exclusively the final global result of the processes of
qualification produced with the considered epistemic referential (G, V )—and
the corresponding genetic class, draws attention upon the absence, in the
classical logic of classes and predicates, of any reference to the epistemic
actions involved by an “object-variable x” or a “predicate P ”. Thereby is
brought into full light the contrast between the active conception on knowl-
edge involved by MRC, and the passive, hypostatizing and absolutizing im-
plications of classical logic.
In a genetic class the undefined, hypostatic shadow-predicates P are
replaced by the views V founded upon aspect-views Vg obeying the definition
D5.1, which consist of effective operations and tests incorporated to a defi-
nite conceptual-operational structure. While the hypostatic object-variable
x is replaced by a definite operation of generation G associated with the
object-entity œG , an operational-conceptual pair (G,œG ) that opens up the
possibility to grasp and to draw up into conceptualization bulks of physi-
cal factuality of which the semantic matter nourishes with meaning all the
levels and sorts of description, the logical one included. Any unnecessary
absolutization is suppressed by the fact that a view is explicitly allowed
to act in the role of generator of object-entity (labelled G(V )) while both
views and generators are allowed to play in other descriptions the role œG
of object-entity. The specificity of the concept of generator introduced by
the definition D4 is not in the least diminished by this absence of strict sol-
220 M. MUGUR-SCHÄCHTER

idarity with the descriptional role G, nor is the specificity of the concept of
a view V as defined by D5.1 and D5.2, diminished by the absence of a strict
solidarity with the role of view. So inside a double-extremity genetic class,
the classical object-variables and predicates—abstract, vague, hypostatic,
absolute, as if out of reach of human action—transmute into a quite definite
and complex operational-conceptual whole of relativized and constructive
epistemic actions.

Outlook on a calculus with double-extremity genetic classes

In what follows we consider exclusively genetic classes involving stable rel-


ative descriptions in the sense of the definitions D14.1, D14.2.1, or D14.3.1
(the genetic classes of testimonial descriptions are too vaguely defined to be
included in a calculus). Furthermore we drop the lower indices as well as the
prefixes and write uniformly C[G.V ].
The logical operations, sum, intersection, complementation, must all
be redefined step by step for the case of genetic classes, in a way fully rela-
tivized to the involved generator G and to the whole content of the acting
view V . The reconstruction requires the definition of laws of composition
of object-entity generators G, of object-entities œG , of views V , and of de-
scriptions D (accordingly to P8 and P10), and it has to be carried out for
all the possible sorts of compositions of genetic classes C[G.V ] (two classes
with both G and V different, or with the same G and different V , or with
different G’s and the same V , or a basic class C[G(o) .V (o) ] and a non-basic
one, or two basic classes, or two non-basic classes (of same order or of differ-
ent orders), or an aspect-class C[G.Vg ]and a general one C[G.V ], etc.). For
instance:
Consider the two genetic classes generated by the successions [G1 .V1 ]
and [G2 .V2 ], both not basic. Then the involved object-entities œG1 and œG2
are conceptual (previously achieved descriptions, or intrinsic metaconceptu-
alizations, or intrinsic models) while the final global results are two descrip-
tions D1 and D2 . Suppose now (G1 ≡ G2 ) = G and V1 6= V2 . Then only one
object-entity œG is involved and the intersection [C[G.V1 ] ∩ C[G.V2 ]] leads
to an (absolutely) void result if V1 and V2 involve no common aspects; while
if V1 and V2 do involve common aspects this intersection yields a description
containing the qualifications present in both D1 and D2 , so one can per-
tinently say that the resulting description is the intersection (or product)
D1 ∩ D2 of D1 and D2 which can be denoted D∩12 . In the same conditions
the union [C[G.V1 ] ∪ C[G.V2 ]] produces a final description that can also be
pertinently called the sum of D1 and D2 and can be denoted D∪12 . Suppose
now on the contrary G1 6= G2 and (V1 ≡ V2 ) = V . Then according to the nu-
cleus of MRC the view V yields a (meta)description of the metaobject-entity
QUANTUM MECHANICS VS. RELATIVIZED CONCEPTUALIZATION 221

œG1 ∪ œG2 where all the qualifications from D1 and all those from D2 are
contained, so one could speak, for instance, of the description of an object-
(2)
sum and introduce the notation D∪œ12 with D∪œ12 ≡ D(2) /G(2) ,œG ,V (2) /
(2)
where G(2) selects the sum-object-entity œG ≡ (œG1 ∪ œG2 ) and V (2) ≡ V .
Etc.

The last example entails that the classical definition of the class of
a predicate P can be progressively approached inside the MRC-logic
by composing additively an increasing number of genetic classes with
distinct generators G of object-entity and identical views V .

In any case the global result of a permitted composition of genetic classes


C[G.V ] is just a relative description. Furthermore, let us consider the logical
voids.

Each class C[G.V ] introduces various semantically relativized voids


tied with corresponding semantically relativized complements:

Given a qualification (gk) from C[G.V ], it introduces its own rela-


tive void—let us label it [∅/(gk)]—that sends to the corresponding relative
complement consisting of the set {(gk)0 }, (gk)0 6= (gk), of all the other
qualifications from C[G.V ]; analogously an aspect-view Vg ∈ V introduces
the relative void [∅/Vg ] that sends to the complement consisting of the set
{Vg0 ], g 0 6= g, of all the other aspect-views from V , so also to the set of all
the qualifications g 0 k from C[G.V ] produced by the aspect-views from V
that are different from Vg . These were examples of relative voids internal
to the genetic class C/[G.V ], i.e., which send to complements contained in
C[G.V ]. If now G and V are regarded as wholes, the genetic class C[G.V ]
introduces three relativized metavoids [∅/G], [∅/V ] and [∅/G, V ] which send
to complements from the outside of C[G.V ], namely to the three exterior
metacomplements with respect to, respectively, G, or V as a whole, or the
referential (G, V ) as a whole (there is no difficulty to characterize these
metacomplements by words).
So a genetic class C[G.V ] introduces a hierarchical organization of
relative voids and of corresponding relative complements sending into def-
inite domains of observation or epistemic action. We are already far from
the connection between a hypostatized “object-variable x” and a hyposta-
tized predicate P —always just a conceptual selector—associated with only
one absolute void. Now, the qualifications gk, the aspect-views Vg , and the
generators G, are all semantic descriptional elements which determine se-
mantic relative voids and the corresponding semantic complements; but, via
222 M. MUGUR-SCHÄCHTER

the symbols that represent them, these semantic relative voids and comple-
ments go into the calculus with genetic classes where they become “logical”
voids and complements that imprint their mark upon a syntax.
We are in presence here of an example in which one can see how
semantic features gain access toward a specific syntactical expres-
sion. What appears on the horizon is a syntax of the extraction and
elaboration of semantic matter, a syntax of conceptualization where
the artificial and illusory frontier between semantics and syntax is
transcended.
The calculus with genetic classes it not yet elaborated, but nothing
hinders to elaborate it. It will have to be worked out in compatibility with
the whole content of the nucleus of MRC. In particular all the restrictions
or methodological rules involved by the frame principle P8, the principle
P10 of individualizing mutual exclusion, and the principle of separation P15
with the concept of relative metadescription D16 entailed by it, will have to
be taken into account systematically. Inside the enlarged framework created
by this calculus, the concept of proof will require reconstruction.
These brief indications suffice for conveying a first notion concerning
the content and the degree of novelty and complexity of the calculus with
genetic classes.

Views of empirical truth. Relative proposition

Consider a genetic class C[G.V ]. It involves as its final global result a cor-
responding relative description D/G,œG ,V /, i.e., some specified structure of
spacetime-gk-values (where one or both frame-aspects of space and time can
be absent), the aspect-index g running over the aspect-views Vg ∈ V that
are different from the frame-aspect VE and VT . Now, following Tarski in this
respect, we note that the mere assertion of the description D/G,œG ,V / is
not itself what is called a proposition. It generates a proposition if and only
if D/G, œG ,V / can be asserted to be empirically true (this is the MRC way
of saying like Tarski that [“the snow is white” is true iff the snow is white]).
Indeed only a previously constituted description can be empirically true or
false. For instance, a basic object-entity cannot exist in the sense of D7 with
respect to an aspect of empirical truth, because it does not exist in the sense
of D7 with respect to any view of comparison (π18.1) while an aspect-view
of empirical truth is an aspect-view of comparison. Indeed it must somehow
compare the mere assertion of the considered description, with some percep-
tions of empirical facts to which this assertion refers; it must somehow be a
view of “verification” able to establish identities or non-identities concern-
ing, on the one hand the assertion, for a definite object-entity, of definite
QUANTUM MECHANICS VS. RELATIVIZED CONCEPTUALIZATION 223

aspect-values gk of definite aspects g, and on the other hand the effective


emergence for that object entity, of precisely those asserted gk-values, when
it is examined via that aspect g. Even not any description can be empirically
true or false (think of the description of a minotaur).
So, quite essentially, each aspect of empirical truth is a meta-aspect
which is relative to an aspect g involved in the description that has to be
“verified”. Like in any identity-valuation, the (two) meta-(aspect)-values of
a meta-aspect of empirical truth, namely “true” or “not true” (false), are
inconceivable in an absolute sense, they can be imagined only relatively to
some definite gk-value of a definite aspect g. If D/G,œG ,V / is an individual
description, then one can desire to establish for each gk-value asserted by
D/G,œG ,V / whether it is true or false; and if D/G,œG ,V / is a probabilistic
description one can desire to establish whether the probabilistic distribution
asserted by it for the values gk of each aspect g is true or false; so one can
also ask: “Is D/G,œG ,V / true with respect to all the gk-values or all the
distributions of gk-values asserted by it?”. But to research a valuation of
empirical truth of D/G,œG ,V / concerning no specified gk-value or distri-
bution of gk-values, would obviously be meaningless. So we introduce the
following definition:

DL.2. Meta[aspect-view] or view of empirical truth. Consider


a meta[aspect-view] consisting of one meta-aspect (τ /g) which is relative
(2)
to an aspect g in the sense of D5.1. Let us designate by [Vτ /g] a corre-
(2)
sponding meta(aspect-view). The meta-aspect (τ /g) from [Vτ /g] is posited
to contain only two aspect-values, namely (τ /g)1 ≡ “true with respect to
g” and (τ /g)2 ≡“false with respect to g”. Accordingly to the general def-
inition D5.1 of an aspect-view—which concerns any aspect of any order –
each meta-aspect (τ )/g must introduce a definite and effective correspond-
ing operation of (τ )/g-examination, as well as an explicit coding rule for
deciding which results of the (τ )/g-examination are to be coded “true with
respect to g” and which ones are to be coded “false with respect to g”.
A meta[aspect-view] of the specified sort will be called a meta[aspect-view]
of relative empirical truth (τ : empirical truth). A view containing two or
more meta[aspect-views] of empirical truth relative to two or more distinct
aspects g will be called a metaview of empirical relative truth and will be
(2)
symbolized by Vτ .

Comment. Consider a previously achieved description D/G,œG ,V /


(2) (2)
and a metaview of relative empirical truth Vτ . If Vg ∈ V and Vτ contains
(2)
a meta-aspect-view [Vτ /g] of empirical truth relative to g that is effective
(2)
with respect to œG , then D/G,œG ,V / and Vτ do mutually exist in the sense
224 M. MUGUR-SCHÄCHTER

(2)
of D7, with respect to g. In this case Vτ is able to qualify the empirical
truth of D/G,œG ,V / with respect to that aspect g. If this is not the case,
then with respect to that g the description D/G,œG ,V / and the metaview
(2) (2)
Vτ do not mutually exist in the sense of D7 and Vτ is not able to qualify
the empirical truth of D/G,œG ,V / with respect to g. If D/G,œG ,V / and
(2)
Vτ do mutually exist in the sense of D7 with respect to all the aspects g
(2)
involved by V , then Vτ can yield for D/G,œG ,V / a complete valuation of
empirical truth.
It follows that according to MRC the concept of empirical truth pos-
sesses some meaning relatively to D/G,œG ,V / only if it is possible to con-
(2)
struct at least one metaview Vτ of empirical truth which exists in the sense
of D7 with respect to D/G,œG ,V /. But this condition is far from being al-
ways insured (as it often seems to be implied). It is a rather restrictive con-
dition, because of the requirements of definiteness, effectivity and codability
entailed by the general definition D5.1 for any (τ /g)-examination. Indeed, of
what can a (τ /g)-examination consist? One possibility is that it consists of
a mere repetition of the Vg -examination itself which inside the genetic class
C[G.V ] leads to this or that aspect-value gk, or this or that probabilistic
distribution of gk-values asserted by D/G,œG ,V , followed by a comparison
between the result obtained in the re-production and the result asserted by
D/G,œG ,V / (the aim of the condition of re-producibility currently imposed
in the “exact” experimental sciences like experimental physics, chemistry,
molecular biology, is precisely to insure possibility of (τ /g)-examinations of
the type specified above). But re-producibility is relatively rare, even for de-
scriptions of physical facts,19 and even for descriptions of physical facts that
belong to what is called an exact natural science. In history, palaeontology,
human biology, police researches, current life, etc., one is in presence of just
testimonial qualifications in the sense of D14.2.2 with respect to which other
sorts of definite, effective and codable (τ /g)-examinations must be invented,
and in many cases this simply is not possible. As for religious, metaphysi-
cal, mythical, poetical testimonial qualifications, the meaninglessness of any
relative metaview of empirical truth is entailed by the very content of the
testimonies.
Consider now a description D/G,œG ,V / for which a complete

19
For instance, if the considered assertion is “yesterday at 14h35, a grain of dust carry-
ing on it a germ X has left my pillow,” it seems highly improbable to be able to construct
for it some meta[aspect-view] of empirical truth founded on reproducibility. So the testi-
monial descriptions are eliminated (which is why I did not call them “descriptions”). In
experimental physics, in chemistry, biology, etc., the specification of metaviews of empir-
ical truth founded upon reproducibility, that be acceptable from all the points of view,
constitutes a basic part of the research.
QUANTUM MECHANICS VS. RELATIVIZED CONCEPTUALIZATION 225

(2)
metaview Vτ of empirical truth has been constructed. Then the valua-
tions of empirical truth of D/G,œG ,V / achieved via the (τ /g)-examinations
(2)
involved by Vτ are in a non removable way relative to these particular
(2)
examinations. In general Vτ is not unique, and with another metaview
(2)
Vτ involving other (τ /g)-examinations one obtains in general other truth-
valuations.
In consequence of the relativizations specified above, the questions of
empirical truth become precise and they admit of definite but only relative
solutions.

This stands in polar opposition to relativism.

The conception on empirical truth exposed above can rather obvi-


ously shown to be in essential agreement with K. Popper’s concept of “rela-
tivity of truth to theory”, as well as with H. Putnam’s views. While Quine,
Kuhn, and many other important thinkers, put less or no accent on the defi-
niteness and effectivity required for a (τ /g)-examination, so in their writings
the question of empirical truth, like that of reference, seems to involve a gen-
eral and irrepressible doom to relativism.
We can now define a relativized concept of proposition:

DL.3. Relative proposition. Consider a description D/G,œG ,V /


(2)
for which it has been possible to construct a complete metaview Vτ of
(2)
relative empirical truth. Consider the metadescription D(2) /G(2) ,œ(2) ,Vτ /
where: the metaobject-entity is œ(2) ≡ D/G,œG ,V / (introduced by a cor-
(2)
responding meta-generator G(2) , namely a conceptual selector); Vτ is
the metaview of empirical truth that exists in the sense of D7 with re-
spect to D/G,œG ,V /, the results of all the involved (τ /g)-examinations
being a priori asserted—tentatively—to consist only of the relative truth-
values “true with respect to the aspect g” for all the Vg ∈ V , which
remains to be validated or invalidated a posteriori by the effective real-
(2)
ization of all the (τ /g)-examinations involved by D(2) /G2) ,œ(2) ,Vτ /. Be-
cause the specified tentative assertion is a “proposition” in the etymological
(2)
sense, D(2) /G2) ,œ(2) ,Vτ / will be called an atomic proposition relative to
(2) (2)
D/G,œG ,V and to Vτ and will be labelled p(D, Vτ ). It can consist either
of the global, integrated formulation “D/G,œG ,V / is true (or false) with
(2)
respect to Vτ ” or of the analyzed set of all the formulations “D/G,œG ,V /
(2)
is true (or false) relatively to Vτ in its assertion concerning that value gk
of that aspect g (or relatively to its assertion of the distribution of gk-values
of g”).
226 M. MUGUR-SCHÄCHTER

Comment. Via the MRC-concepts of metaview of empirical truth


and of relative proposition, the calculus of genetic classes C[G.V ] leads to a
corresponding relativized calculus of propositions, where the truth-value of
the final description produced by a composition of genetic classes C[G.V ] has
to be established as a function of: the nature of the composition; the involved
metaviews of empirical truth; the values of empirical truth assigned via these
to the descriptions produced by the classes involved in the considered com-
position of classes. So, while in the classical approach a truth-valuation is
from the beginning on involved in the definition of the class of a predi-
cate P, in the MRC genetic logic the genetic classes are clearly separated
from the corresponding propositions, of which the truth-valuations require
different, explicit, analyzed, non-trivial relative specifications. This, at first
sight, might seem to be a huge complication, to be avoided at any price.
But in fact it is a complexification of the treatment that can determine with
any desired precision the configuration of the channels along which semantic
matter is adduced into logical syntax. Too often, for the sake of simplicity,
false absolutes are introduced, which block the growth of thought.

A non-classical logical stratum concerning strictly singular phys-


ical factuality

Consider a basic genetic class C[G(o) .V (o) ]. Even if the basic description
D(o) /G(o) ,œ(o) ,V (o) / involved by this class is called “individual”, in conse-
quence of the condition of stability from the general definition D14.1 and its
particularization D14.3.1, it involves nevertheless a big number of repetitions
of the realization of each succession [G(o) .V (o) ].
The epistemic action leading to D(o) /G(o) ,œ(o) ,V (o) / as a whole, no
matter whether D(o) is individual or probabilistic, is always directly
placed on the level of statistics.
However, by penetrating inside D(o) and taking into account only
(o) (o)
two distinct successions [G(o) .Vg1 ] and [G(o) .Vg2 ], it is possible, by use of
the concept D14.2.2 of testimonial description, to dig down to the level of
the strictly individual qualifications, and to define for these a semantical
character which determines a certain corresponding logical form. This is an
innovation with respect to classical logic. We proceed as follows.
(o) (o)
Consider two distinct successions [G(o) .Vg ] and [G(o) · Vg0 ]0 that
0 0
have been realized either with g ≡ g or with g 6= g, no matter, but have led
to two different aspect-values gk 6= gk 0 or gk 6= g 0 k 0 , respectively. These two
successions—with their outcomes included—do not insure a test of descrip-
tional stability as required by D14.1 or D14.3.1. So they are not descriptions
in the sense of the mentioned definitions, they are just two testimonies in
QUANTUM MECHANICS VS. RELATIVIZED CONCEPTUALIZATION 227

the sense of D14.2.2, say θ1 and θ2 . Now, because θ1 and θ2 involve by hy-
pothesis two distinct registered aspect-values, each one of these testimonial
descriptions required its own realization of a replica of the object-entity œ(o)
(o)
produced by G(o) . So, if we label œj one given realization of a replica of
(o) (o)
the basic object-entity œ(o) , the two testimonies θ1 (œj ) and θ2 (œj ) are
mutually incompatible because they cannot both realize for that one given
(0)
replica œj .

(o)
As soon as a restriction to only one definite replica œj of a basic
object-entity œ(o) is posited—not only restriction to no matter how
many replicas of one sort of basic object-entity œ(o) as defined by
a given operation G(o) , but furthermore restriction also to only one
replica of that sort of basic object-entity—there arises a mutual in-
(o)
compatibility between the factual realizability of θ1 (œj ) and that of
(o)
θ2 (œj ). This holds even if the qualifications involved by θ1 and θ2
concern both one same basic aspect g.

This is a mutual exclusion of a semantical nature. But via the concept


of empirical truth it entails a formal logical consequence. To show this we
proceed as follows. To begin with, we define:

DL.4. Basic relative atomic testimonial proposition. We call


basic relative testimonial proposition and we label p(θ(o) ) the tentative as-
(o)
sertion of the empirical truth of a relative basic testimony θ/G(o) ,œ(o) , Vg /
(2)
(with respect to some definite view of empirical truth Vτ /g (o) supposed to
have been constructed); which tentative assertion remains to be validated
(2)
or invalidated via the (τ /g (o) )-examinations involved by [Vτ /g (o) ].

Comment. A basic aspect of empirical truth concerning a basic


(o)
testimony θ/G(o) ,œ(o) ,Vg / can consist, for instance, of the consensus con-
cerning the genesis and the outcome of the testimony θ, among an arbitrar-
ily big number of observers that have watched and witnessed together these
non-repeatable phenomena.
(o) (o) (o) (o)
So to the two testimonies (θ1 ) = θ1 (œj ) and θ2 = θ2 (œj ) there
(o) (o)
correspond two testimonial propositions p1 [θ1 (œj )] and p2 [θ2 (œj )]. Now
(o) (o)
since θ1 (œj ) and θ2 (œj ) cannot be both realized because they involve by
(o)
hypothesis two different outcomes for one same replica œj of the involved
(o) (o)
basic object-entity œ(o) , a fortiori p1 [θ1 (œj )] and p2 [θ2 (œj )] cannot be
both true. So:
228 M. MUGUR-SCHÄCHTER

(o) (o)
A logical conjunction of p[θ1 (œj 0] and p[θ2 (œj )] is devoid of factual
counterpart. It cannot be defined, which is a case different from that
in which it can be defined but comes out to be false.

This can be better understood by the help of truth-tables: Given two


propositions p and q, their logical product p ∧ q is defined by:

p q p∧q
T T T
T F F
F T F
F F F

(o) (o)
What happens if p ≡ p1 [θ1 (œj )] and q ≡ p2 [θ1 (œj )]? In this case the
top line “TTT” represents a combination which, factually, is systematically
impossible. The factually possible cases are only

p q p∧q
T F F
F T F
F F F

But what this last set of possibilities claims, is that the logical product
p ∧ q simply does not “exist”, factually, since it never is factually
true.20
20
Wittgenstein [25] made an analogous analysis related to another sort of factual mutual
spacetime exclusion: “I have said elsewhere that a proposition ‘reaches up to reality,’ and
by this I meant that the forms of the entities are contained in the form of the proposition
which is about these entities . . . . For the sentence, together with the mode of projection
which projects reality into the sentence, determines the logical form of the entities . . . . For
if the proposition contains the form of an entity which it is about, then it is possible that
two propositions should collide in this very form. The propositions “Brown now sits in this
chair” and “Jones now sits in this chair” each, in a sense, try to set their subject term on the
chair. But the logical product of these propositions will put them both there at once, and
this leads to a collision, a mutual exclusion of these terms . . . . It is, of course, a deficiency of
our notation that it does not prevent the formation of such nonsensical constructions, and
a perfect notation will have to exclude such structures by definite rules of syntax. These
will have to tell us that in the case of certain kinds of atomic propositions described in
terms of definite symbolic features certain combinations of the T’s and F’s must be left out
(T: true; F: false). Such rules, however, cannot be laid down until we have actually reached
the ultimate analysis of the phenomena in question. This, as we all know, has not yet been
achieved“. Wittgenstein’s propositions “Brown now sits in this chair” and “Jones now sits
in this chair” are related with a dual spacetime mutual exclusion (two distinct sorts of
object-entities are involved, not only one) and furthermore a spacetime mutual exclusion
QUANTUM MECHANICS VS. RELATIVIZED CONCEPTUALIZATION 229

It claims this in the amputating “purely syntactical” language of clas-


sical logic. But what is thus claimed is not a purely syntactical matter, it is a
(o)
matter of syntax which directly expresses a matter of fact. If p ≡ p1 [θ1 (œj )]
(o)
and q ≡ p2 [θ2 (œj )], the logical product p ∧ q considered above is mean-
ingless with respect to the value “true” of any aspect g of any constructible
view of empirical truth with respect to which both p or/and q do exist in
the sense of D7. This is so in consequence, not of the falsity of either p or q
considered separately, but in consequence of the fact prior to such a falsity,
(o)
that the realizability of the testimony θ1 (œj ) is incompatible with that of
(o)
the testimony θ2 (œj ), so that p and q cannot coexist. To represent this
new sort of situation by still saying in an inertial and non specific way that
p ∧ q is “false”—exactly as we say in the cases when p and q can coexist
but one of them is false—amounts to a too loose formalization-and-language
which by construction is unable to express the specificities of a whole def-
inite category of cases. Obviously the aim of maximal formal “generality”
cannot justify such a categorial non-specificity. In a well-adjusted logical for-
malization the situation from the last table requires an own syntactical sign
that shall prevent void writings of logical products p ∧ q that are a priori
impossible factually.
This is the usually so fuzzily understood core of what is called “quan-
tum logic”, reflected there in such a truncated and distorting fashion.21
But as soon as two or more replicas of a given object-entity are al-
lowed (so a fortiori if also two or more sorts of object-entities are allowed) the
mutual exclusions founded on the unicity of the involved replica of object-
entity vanish, and a factual counterpart can be defined for the logical con-

that can happen or not (if in the second proposition, instead of Jones, we set “Brown’s
bacterian flora” there is no exclusion any more). Therefore this kind of dual spacetime
mutual exclusion cannot be expressed by a principle like P10. But it is very striking indeed
that—without benefiting of guidance by quantum mechanics, which in the present work
led toward “the ultimate analysis of the phenomena in question”—Wittgenstein as early as
1929 identified the decisive individualizing role played by spacetime in the factual mutual
exclusions of two propositions (he labels propositions like those considered above by the
group of letters PT where P means place and T means time: our condition of one same
replica of the object-entity œ(o) corresponding to G(o) amounts to a labeling by the same
values PT of space-and-time!). It is also striking that, notwithstanding Wittgenstein’s
work quoted above, the illusory belief of independence of syntax, on semantics, still is so
strong up to this very day.
21
In quantum mechanics the distinction between the individual level of description
and the statistical one is not sufficiently clear, so the ways of speaking often seem to
involve that qualifications by two mutually incompatible observables are always mutually
exclusive, while qualifications by two compatible observables are never mutually exclusive:
the decisive role of—exclusively—the restriction, or not, to only one replica of the
(o)
involved object-entity œj , is not recognized.
230 M. MUGUR-SCHÄCHTER

junction of any two successions [G(o) .V (o) ], even if they correspond to mutu-
ally incompatible basic views. Then, however, one finds oneself already in
the realm of statistics, and there, grosso modo, the “Boolean” logic, so the
algebras from the classical probability spaces, do operate (cf. note 21).

The classical assumption of a non-restricted possibility of logical con-


junction presupposes statisticity. The classical Boolean logic is quasi
systematically statistical. It overlooks the specificities of strict indi-
viduality.

By its “universals” (at least) classical logic usually begins above the
level of strict individuality and then keeps floating over it, loose and dead,
cut away from its unknown roots implanted in strict factual individuality.
While only a level of logical conceptualization where strict individuality is
explicitly characterized can contain a common foundation for classical logic
and classical probabilities (cf. 5.2).
For the particular case of quantum mechanics [13], I have already
introduced a logical conjunction restricted by a syntactical sign of factual
mutual exclusion between two propositions reflecting the unicity of the in-
volved replica of object-entity. This permits to deal with the question of
quantum logic in a much deeper way than the usual one. Now, the mentioned
(o)
approach can be generalized to any two testimonial propositions θ1 (œj )
(o)
and θ2 (œj ). When this is done it becomes possible to effectively construct
an MRC-calculus with testimonial propositions which connects the level of
strict factual individuality, with the statistical level of logic, via a very first
stratum of logical form where the conjunction is not universally permitted.

Like the relativization to semantic features of the syntactical logi-


cal void, the dependence of the domain of pertinence of the logical
conjunction, on semantical features (the mutual incompatibility of
(o) (o)
two testimonial propositions θ1 (œj ) and θ2 (œj )) and so the mu-
(o)
tual factual exclusion of the corresponding propositions p1 [θ1 (œj )
(o)
and p2 [θ2 (œj ), illustrates again how factuality, semantics, can de-
termine logical form.

The MRC-status of the “objects” of the classical logic of classes


and predicates

Inside the general category of genetic classes, the classical concept of class
is re-obtained in only the following two cases:
QUANTUM MECHANICS VS. RELATIVIZED CONCEPTUALIZATION 231

(a) A basic genetic class of the type C[G(o) (V (o) ).V (o) ] is involved,
where V (o) denotes a human biological sensorial view. In this case the gen-
erator G(o) (V (o) )—i.e., V (o) itself but in the role of generator of object-
entity—even though it is basic, is not explicitly perceived to create out
of the physical reality the corresponding object-entity, namely the field of
sensitivity of V (o) ; while the basic view V (o) , again the view involved in the
description but which now also plays the role of a view, can be assumed with-
out inner contradiction to qualify the created object-entity without changing
it. This particular sort of basic referential produces a very simplified ver-
sion of basic description D(o) that can be, and indeed is, spontaneously
metaconceptualized intrinsically, by an implicit process; and then it further-
more is immediately reduced implicitly to the corresponding intrinsic model
(1)
M (œ(o) )/[V (o) , VI ], where the relativities to the basic view V (o) and to
(1)
the intrisizing view VI remain hidden, only the model M (œ(o) ) itself is
perceived, and so it is taken to be absolute. This model is what is illusively
felt to somehow exist eternally and immutably, independently of any ob-
server, in an abstract Platonian space where it stays available for passive
perceptions of [truth’s of P ’s] (cf. D19.1, and D19.2 with their comments,
and Sec. 5.1.1). This—the models M (œ(o) ) cut from their relativizing ties
with the basic transferred descriptions wherefrom they stem—is the basis of
the Platonian realism (in the scholastic sense), which down to the present
day grasps the minds with irresistible force. The logicians and mathemati-
cians are particularly exposed to this force because they have found methods
to distil consistent systems of very abstract models M (œ(o) ) which are so
perfect that a posteriori they seem to be endowed with divine pre-existence
and supreme intelligibility (think of Pean’s arithmetic).
(b) A basic genetic class of the type C[G(o) (V (o) ).V (o) ] is involved
where V (o) denotes extensions by apparatuses of the domain of human bi-
ological sensorial aspect-views. All the preceding remarks are valid for this
case also. The intrinsic models elaborated in this somewhat enlarged frame-
work belong to the realm of exact classical sciences (think of what is called
atomic spectra, and the corresponding intrinsic models of atoms) to most of
which the classical logic still applies.
In both cases mentioned above the content of the epistemic operator
playing the role G, identifies with the content of the view V which plays
the role of a view, and furthermore this view V is reduced to an undefined
and structureless abstract “predicate P ”. So G ≡ V ≡ P , all the involved
descriptional actors being identified to P . This point-like degeneration is
what entails the loss of awareness of the ineluctable action, in any description
and so in any proposition, of also a generator G of object-entity. Correlatively
the “direction of conceptualization” defined by a double-extremity genetic
232 M. MUGUR-SCHÄCHTER

class C[G.V ], gets lost also. The classical definition of a class determined by
(the truth of) exclusively a predicate P is just tangential to the superficial
level of the already verbalized-conceptualized intrinsic models represented by
“object-variables x”, a definition which is loose like the needle of a compass
on the surface of the earth.
So we have recovered here in analyzed terms a conclusion already
asserted in the preliminary critical comments from Sec. 5.1.1:

Inside MRC, the domain of “objects” directly considered in the


classical logic is found to consist of exclusively intrinsic models
(1)
M (œ(o) )/[V (o) , VI ], always conceptual constructs extracted from
spontaneously achieved implicit intrinsic metaconceptualizations of
degenerate transferred descriptions D(o) produced by successions of
the particular type [G(o) (V (o) ).V (o) ] where the human biological ap-
paratuses cumulate the role of generator of object-entity and the role
of view.

These “objects” never disclose the bulk of a as yet non-conceptualized


physical factuality from their cores, wherefrom any conceptualization stems
via basic transferred descriptions. The connection between an intrin-
(1)
sic model M (œ(o) )/[V (o) , VI ] and the corresponding basic description
(o)
D(o) /G(o) ,œG ,V (o) / remains ignored because both the basic view V (o) and
(1)
the intrinsizing metaview VI are wired into the morphology and the reflex
(1)
functioning of our bodies, so the relativities to this pair of views (V (o) ,VI )
remain hidden to the immediate natural perception of the human mind. The
neurobiologists and the cognitivists are now studying them intensively from
a psycho-biological standpoint. But among the sciences of non-biological do-
mains of the physical reality, only quantum mechanics has succeeded to get
down to these cores of a-conceptual physical factuality hidden inside the
classical models, and it has represented their extraction as well as their very
first transposition in communicable terms, by basic transferred descriptions.
It has represented all this indeed, but only cryptically, mathematically from
the start on, and without being able to formulate their descriptional status,
nor to accomplish also the subsequent descriptional phase of intrinsic meta-
conceptualization. The integral conceptual trajectory that leads from the
basic transferred descriptions to classical models M (œ(o) ), remained hidden
to quantum mechanics also, and in consequence of this the universal sig-
nificance of quantum mechanics itself remained hidden. So the possibility—
always—of at least a minimal intrinsic model in the sense of D19.3, has not
been pointed out, and the universal rooting of any model, in physical fac-
tuality, remained non perceived. And now, when eventually all this becomes
QUANTUM MECHANICS VS. RELATIVIZED CONCEPTUALIZATION 233

apparent and so a general law of growth of the processes of conceptualiza-


tion is brought forth, it will be tried for some time, no doubt, to ignore or
even to deny it, because a positivistic philosophy has had time to constitute
and to consolidate itself, and so now it opposes its own inertial resistance.
In sum, in classical logic we circulate swiftly on an aerial net of smooth
highways for deduction, erected out of models drawn from a thick stratum of
unsuspected hidden conceptualization that keeps us far from the nourishing
background of an as yet non-conceptualized physical reality. The MRC ge-
netic logical approach explicates the presence of this stratum and its whole
morpho-functional structure into which the classical models M (œ(o) ) are
fixed by innumerable genetic threads. This offers now this stratum, as well
as the models, to control and deliberate use and also to confrontation with
the neural biological processes.

Formal systems versus genetic classes

It might now seem that the integral domain of the MRC-logic of double-
extremity genetic classes, can be obtained by simply adding to the sub-
domain corresponding to classical logic as specified above, the domain of
basic double-extremity genetic classes C[G(o) .V (o) ] with G(o) 6= V (o) and
where in general V (o) changes to a significant degree the object-entity cre-
ated by G(o) . But in fact such a juxtaposition would not exhaust the domain
of the genetic classes C[G.V ]. Indeed it would leave out all the double-
extremity genetic classes produced by a conceptual epistemic referential (de-
generate or not) that is creative and yields stable relative descriptions of type
D14.2.1. While the natural representations, and even the scientific ones, quite
currently do involve double-extremity creative conceptual genetic classes,
notwithstanding that classical logic does not define them.
This is a paradoxical situation of which a massive illustration can
be found in mathematics as well as in the modern formal logic itself ! The
central concept in these disciplines is that of a formal system S. A finite
formal system consists of a finite list of primitive symbols, a finite list of
terms formed with primitive symbols, a finite list of well-formed expressions,
a sub-set of well-formed expressions called axioms, and a finite list of rules
of transformation of a given well-formed expression, in another one. In a non
finite formal system the list of primitive symbols can be indefinitely enlarged
(as in Peano’s arithmetic). The well-known concept of formal system needs
no further specification in order to be reconsidered inside MRC, so we do not
introduce a specifically MRC-definition. Let us simply note that a formal
system is generated by the conceptor’s mind via a generator of object-entity,
say GS , that consists of an epistemic action upon the zone of “reality” (in
the sense of D2) consisting of the “conceptual reality” from the conceptor’s
234 M. MUGUR-SCHÄCHTER

mind, his knowledge included, say RC (cf. D4). The process of generation
of this object-entity is quite essentially creative.
We now try to specify what a formal description is accordingly to
MRC.
Once constructed, a formal system S can be regarded as the abstract
zone or domain from “reality” in the sense of D4 where all the formal de-
scriptions permitted by S are carried out. It is a sort of conceptual platform,
smooth, stable and solid, conceived in order to permit us to achieve on it par-
ticularly precise descriptional trajectories. So S itself has to be constructed
in the first place; afterward one can elaborate also descriptions “in” S. This
preliminary condition for the achievement of a formal description will have
to be somehow explicitly expressed in the specification of the notations that
characterize a description in S.
In all the mathematical or logical treatments it is assumed more or
less implicitly that as soon as the formal system S is given, ipso facto one
knows how to work with it because the rules are incorporated. But inside
MRC one is obliged by method to always specify explicitly the epistemic
referential (G, V ) inside which a (relative) description D/G,œG ,V / is at-
tempted, as well as the involved object-entity œG . So we ask: of what does
a formal description consist, what plays in it the role of generator of object-
entity, what plays the role of object-entity, and what plays the role of view?
A formal description from a given formal system S is a finite proof
carried out inside S. Let us call it here a proof-description and label it DjS
where j is an index that distinguishes between the various proof-descriptions
from S. By classical definition, a proof-description DjS consists of a finite
sequence of n well-formed expressions that are all permitted in S in con-
sequence of the fact that the sequence always starts with an axiom or a
well-formed expression known to follow from the axioms from S (theorem)
which then, in a sequence of n descriptional steps Djk S , k = 1, 2, . . . , n, is

progressively transformed by the combined use of rules of transformation


from S and of “lateral” introductions of other axioms from S or of already
proven theorems from S, the end of the sequence being reached when an “in-
teresting” well-formed expression emerges which previously was not known
to follow from the axioms of S and which now is listed as a new theorem in
S.22
Let us denote by œj this final well-formed expression from S: it can
be regarded—at least a posteriori—as the object-entity of the considered
22
Since a given theorem is the result of a definite proof, it might seem inconvenient
to mix the definitions of distinct proofs by making use in the definition of a proof A, of
theorems established in other proofs. But the use in A of a theorem established in another
proof B is just a short-hand for the—equivalent—introduction of that whole other proof
B. So the definitions of the various proofs are separable.
QUANTUM MECHANICS VS. RELATIVIZED CONCEPTUALIZATION 235

proof-description DjS .
So the involved generator of object-entity is by definition that which
generates œj . For this the generator must dispose of S. Therefore it is per-
tinent to posit for the notation of the generator of œj the form of a product
of two successive operations of generation, say GSj = Gj GS where: GS acts
first on the zone RC from reality consisting of the conceptor’s mind, thereby
producing the zone RS of “reality” consisting of the formal system S; Gj
acts subsequently, on RS ≡ S, thereby producing the well-formed expres-
sion œj to be proven. (Of course this analysis is only notational. Once S
has been created by the epistemic action labelled GS it remains indefinitely
available, and there is no need to effectively re-produce its generation for
each object-entity œj : only Gj has to be chosen and acted with in each
case.)
The aim of DjS is to establish whether yes or not œj is provable
inside S. So œj has to be examined by a formal view of provability inside
S. (Retroactively it is always possible to represent the proof-process in this
way, though in fact most often œj emerges constructively together with its
proof). We introduce now an explicit MRC-definition of the view that acts
in a proof-description:

DL.5.View of demonstrability in S.23 Consider a formal system


S. From the classical definitions of S and of the proof-descriptions DjS from
S it follows that the view which acts in DjS must be able to qualify the
object-entity œj in terms of the aspect-values of two aspects g1 and g2 ,
namely: (a) an aspect of form inside S, g1 ≡ Φ, endowed with two aspect-
values, say, respectively, Φ-yes (well-formed inside S),Φ-no (not well-formed
inside S); and (b) an aspect of transformation inside S, say g2 ≡ Θ, equally
endowed with two aspect-values, say Θ-yes (correctly transformed inside
S), Θ-no (not correctly transformed inside S). The view consisting of these
two aspects is a formal view relative to S that will be called a view of
demonstrability in S. It will be labelled VdS .
Comment. The upper index S stresses that the formal view VdS
is extracted from S (remember that according to MRC this dependence
between VdS and S, so also between VdS and GS (which is involved as a
“factor” in the global generator GSj = Gj GS ) is a restriction with respect to
the most general situation of mutual in-dependence between the generation
operators and the acting view). The aspect Φ from VdS qualifies accordingly
to the list of well-formed expressions posited in S, and its aspect Θ qualifies

23
We choose the word demonstrability only in order to index by d: the word provabil-
ity would require the index p that might lead to confusion with indexes concerning the
concepts of proposition or of probability.
236 M. MUGUR-SCHÄCHTER

accordingly to the transformation-rules posited in S. So in fact what VdS


is able to ascertain for any expression from a proof-chain, is just that it is
formally consistent with the requirements of well-formedness and of ways of
transformation from S.

VdS is—exclusively—a yes-no filter concerning well-formedness in S


and transformation in S. Nothing else.

The fundamental but often obscure problems concerning the relations


between demonstrability in S and “truth”, will be discussed in the next
paragraph. For the moment, in what follows immediately we speak only of
demonstrability.
So the epistemic referential corresponding to DjS is (GSj , VdS ) ≡
(Gj GS , VdS ).
Now, how can we represent the emergence of a proof-description
DjS /GSj , œj , VdS /? Can it be conceived as the result directly produced by
a corresponding genetic class C[GSj .VdS ], i.e., as the result of re-productions
of a set of successions [GSj .VdS ] defined from the start on? The structure
posited for a proof-description DjS shows immediately that the answer is
negative. Indeed in the course of the elaboration of DjS the view VdS does
not work constantly on one same object-entity, namely the object-entity œj
generated by GSj which has to be proven in S. VdS works on other inter-
mediary well-formed expressions produced by other generators that become
possible progressively while the proof-description DjS is developing. So a
more analyzed answer is needed here. It can be established as follows.
We have noted before that the integral description DjS emerges by n
successive mutually different descriptional steps. Let us denote by Djk S , k =

1, 2, . . . , n the k-th step. This is a one-step “elementary” proof-description


S /GS , œ , V S /. It involves a generator of object-entity GS = G G
Djk jk jk d jk jk S
S
which is different from the generator Gj = Gj GS from the epistemic ref-
erential (GSj , VdS ) corresponding to the integral proof DjS and produces a
“local” object-entity œjk that is different from the object-entity œj from
DjS /GSj , œj , VdS / (the index j which has been conserved in the notations
Djk and œjk reminds that the designata of these notations are both referred
to the object-entity j). In the first descriptional step Dj1 S /GS , œ , V S / the
j1 j1 d
object-entity j1 with which Dj1 S ends (which had to be ascertained) is pro-

duced by a generator GSj1 that still acted on the zone from the conceptual
reality consisting of S itself, like in the case of the global generator GSj from
the integral proof-description DjS , but nevertheless this is already another
generator because it produces the object-entity œj1 6= œj . And for k > 1 the
corresponding generator GSjk does not even work on S any more. It works
QUANTUM MECHANICS VS. RELATIVIZED CONCEPTUALIZATION 237

on [S ∪ {œjm }], m = 1, 2, . . . , k −1 where {œjm } is the set of well-formed


expressions of which the demonstrability in S has been established by the
sequence Dj1 S DS , . . . , DS
j2 jk−1 of the previously accomplished elementary de-
scriptional steps. (The set {œjm }, m = 1, 2, . . . , k−1 has to be added to S
because now it is explicitly available in the conceptor’s mind and the k-th
choice of an object-entity œjk takes support on this set also, not only on
S any more). So the integral proof-description DjS emerges by an “addi-
tive composition in succession” of the n elementary step-descriptions Djk S ,

k = 1, 2, . . . , n where the generator of object-entity and the corresponding


object-entity change—in a way that is not prescribed by S – while the view
remains the same. In these conditions we can show that:

ΠL6. Proposition. A proof-description DjS can be considered to be


produced by a non-degenerate double-extremity and creative genetic class.

“Proof”. A one-step description Djk S , for any k between 1 and n,

can be non trivially regarded as the result of the corresponding double-


extremity genetic class C[GSjk .VdS ]. Indeed the succession of epistemic op-
erations [GSjk .VdS ] is indefinitely repeatable and its result stays unchanged,
namely it is the k-th final well-formed expression œjk that has been shown
to be provable. So we are strictly in agreement with the concept of an
individual conceptual description produced by a double-extremity genetic
class, as formed by the definitions D14.1, D14.2.1, and DL1.2. Further-
more, the ordered juxtaposition in succession of the elementary proof-
descriptions Djk S brought forth by the elementary genetic classes C[GS .V S ]
jk d
with k = 1, 2, . . . , n, yields a definite new description in the sense of D14.2.1,
namely precisely the integral proof-description DjS as defined from the start
on. So we can write
X
DjS /GSj , œj , VdS / ≡ S
Djk /GSjk , œjk , VdS /,
k

k = 1, 2, . . . , n where in the last descriptional step Djn S /GS , œ , V S / the


jn jn d
object-entity œjn has the same content as œj but is generated by the gen-
erator GSjn —heuristically different from GSj —which acts on [S ∪ {œjm }],
m = 1, 2, . . . , n−1, not exclusively on S like the (in general fictitious) gener-
ator GS . So DS /GSj , œj , VdS / is the global result of the “sum in succession”
P Sj S j S
k Djk |Gjk , œjk , Vd | (see “Outlook on a calculus with genetic classes”).
S · V S ] ∼ [GS · V S ]. In this sense
P
Which means that we can write k [Cjk d = j d
Dj can indeed be considered to be produced by the genetic class [GSj · VdS ]
S

(which, globally, can be repeated P an arbitrary number of times, once it has


been obtained from the sum k [GSjk · VdS ], k = 1, 2, . . . , n. This establishes
238 M. MUGUR-SCHÄCHTER

πL6.24

Comment. In the first place, the fact that œjn has the same content
as œj while on the other hand GSjn 6= GSj , might seem to contradict the one-
one relation G − œG posited in D4. But in fact GSjn working on [S ∪ {œjm }],
m = 1, 2, . . . , n − 1 amounts to an effective and explicit representation of
precisely the global generator GSj , iff the proof of œj succeeds (if not, the
very concept of what is denoted GSj is discarded). So the effective expression
of GSj is given by the definition
X
DjS /GSj , œj , VdS / ≡ S
Djk /GSjk , œjk , VdS /, k = 1, 2.....n.
k

What appears here is that, as already remarked, the a priori operation of


generation GSj of the well-formed expression œj to be proven is in general just
an a posteriori fiction, that in fact œj is obtained progressively, construc-
tively, by trial and error, while DjS /GSj ,œj ,VdS / ≡
P S S , œ ,V S ,
D
k jk /G jk jk d
k = 1, 2, . . . , n, is being sedimented. And once œj and Gj have been S

settled—together—the one-one relation between them is insured: I postu-


late that two different proofs never have identical results, they can imply
the same result, but each one also has specific entailments.
In the second place, the definition
X
DjS /GSj , œj , VdS / ≡ S
Djk /GSjk , œjk , VdS /, k = 1, 2, . . . , n,
k

obtained above suggests that the to-be-established calculus with genetic


classes will include a general definition of “additive composition in succes-
sion” of certain types of genetic classes. (Such a definition can appear to be
important in an attempt at a mathematical formalization of MRC).
In the third place:
24
It is noteworthy that in a certain sense the structure found for the process of emer-
gence of a proof-description presents certain similitudes with the way in which the basic
transferred quantum mechanical description of a microstate is brought forth. Indeed the
quantum mechanical measurement-evolutions draw into the realm of the observable and
communicable, aspects of the studied microstate that can be conceived a posteriori as rela-
tive potentialities possessed ab initio by the studied microstate which have been actualized
by the measurement evolutions. While the provability in S of the studied well-formed ex-
pression œj can also be conceived a posteriori as a potentiality of S to yield œj that has
been actualized by the proof-description DjS /GS S
j ,œj ,Vd /. The visibility of all the interme-
S
diary steps Djk —devoid of equivalent in the quantum mechanical case—stems from the
fact that here the cognitive situation is different, the object-entity as well as the whole
descriptional process being conceptual, which permits a uniform perceptibility that cannot
be realized for a physical microprocess.
QUANTUM MECHANICS VS. RELATIVIZED CONCEPTUALIZATION 239

While a formal system S itself is an object-entity generated by a


creative abstract generator, i.e., it does not pre-exist like a “value”
of a classical “object-variable x”, furthermore the concept of formal
proof inside S, in its turn, appears to have the nature of a relative
description produced by genetic classes, not by pre-existing shadow-
“predicates P”: the concepts that are the very core of the modern
classical logic stem from epistemic actions that are not defined inside
modern classical logic.

This paradoxical situation illustrates strikingly how we currently act


inside conceptual volumes that are not included in our explicit representa-
tions.
We sum up. The whole set of the researched MRC-terms concern-
ing a proof-description is this. The object-entity generator GSj = Gj GS is
a fundamentally creative conceptual generator consisting of, first the con-
struction of the stable formal “ground” consisting of S itself, and then, out
of S, of the choice or the construction of the object-entity œj to be proven
in S. The view is the formal view VdS of demonstrability in S, extracted
from S, so a view that depends on S (or, equivalently, on GS , so on GSj ).
So the epistemic referential where any proof-description DjS is achieved is
(GSj , VdS ). The explicit structure of a proof-description in S is
X
DjS /GSj , œj , VdS / ≡ S
Djk /GSjk , œjk , VdS /,
k
k = 1, 2, . . . , n, œjn ≡ œj , GSjn 6= GSj .

The MRC-relations between empirical truth and demonstrability

We have shown that a view of demonstrability VdS in the sense of DL.5 has
nothing to do with empirical truth as ascertained by a metaview DL.2. How-
ever it is quite currently said of an expression which has been proven in S
via VdS that it has been shown to be “true”. Those who want to be more
specific make sometimes use of the expression “formally true” in S. Further-
more a well-formed expression that has been proven in S, is often referred
to as a “proposition” which, “because” it has been proven in S, necessarily
is also “true”, not in S this time, just true in the sense of “empirical math-
ematical truth”. Whereas inside the MRC-logic a (relative) proposition in
the sense of DL.3 is a concept quite different from a well-formed expression
from a formal system (which is consistent with the axioms from S via the
transformation rules from S), it concerns empirical truth, not consistency.
In fact all the formulations of the sort mentioned above, where the word true
is made use of, are related with the supposition that the axioms from S are
240 M. MUGUR-SCHÄCHTER

empirically true. If this is not made clear it might entail much confusion.
So below I shall now explicate the MRC-relations between propositions tied
with empirical truth, axioms, and demonstrability inside a formal system.
For the logicians and mathematicians such specification are certainly trivial:
I apologize for this.
(2)
According to MRC, a view Vτ of empirical truth (DL.2) is a
metaview which can exist in the sense of D7 only with respect to a previously
achieved relative description D/G,œG ,V /. Only a piece of meaning that has
been elaborated into a relative description previously, independently of any
question of truth, can afterward be found to be empirically true or false;
and this, if it can happen at all, can happen only with respect to a specified
(2)
metaview Vτ of empirical truth. An absolute assertion of empirical truth is
rejected inside MRC as devoid of significance. This is why the MRC-concept
(2)
of proposition p(D, Vτ ) defined in DL.3 is a metaconcept, and is doubly
relative. It must involve an independently constructed description and a
metaview of empirical truth constructed for definite aspects.
A view VdS of demonstrability in a formal system S (DL.5) can exist
in the sense of D7 only with respect to well-formed expressions from S.
Though inside another formal system which is a metasystem with respect
to S it might also be possible to construct a metaview of demonstrability in
S, VdS is not quintessentially a metaview.
Now, in general a relative description D/G,œG ,V / is not a well-
formed expression from a formal system S, so in general it does not exist
in the sense of D7 with respect to a view of demonstrability VdS ; in gen-
eral well-formedness and correction of transformation inside some formal
system have no relevance with respect to a relative description. And vice
versa, in general a well-formed expression from S is not a relative descrip-
tion, it is just a sequence of signs permitted inside S, expressly posited to
have been purified of any semantics, of any meaning; it is by construction
“invisible” to the views of empirical truth which consist of procedures for
testing assertions of values of empirically perceivable aspects g. So the rel-
ative descriptions, the metaviews of empirical truth and the propositions,
form a group of essentially semantical concepts which simply have nothing
to do with the well-formed expressions and the view of demonstrability from
a “purely” formal system.
In these conditions, what is the reason why provability and truth are
so readily coalesced with one another?
The main reason is the current assertion that the axioms from a
formal system S are posited to be true. But in fact no formal system at
all is—stricto sensu—concerned by this way of speaking. The axioms are
posited to be true only in the interpretations of a formal system S, if these
QUANTUM MECHANICS VS. RELATIVIZED CONCEPTUALIZATION 241

exist, or in the deliberate formalizations of this or that theory of a domain of


empirical facts (physical or abstract) that has first been constructed quite
independently of any formal system and afterward has been axiomatized,
and then formalized. In both these cases the “axioms” are well-formed ex-
pressions from the formal system obtained in this way, which are explicitly
(2)
constructed so as to translate relative propositions p(D, Vτ ) posited to ob-
tain the empirical-truth-value “true” when the description D involved by
(2) (2)
p(D, Vτ ) is examined via the metaview Vτ of empirical truth relatively
(2)
to which p(D, Vτ ) is defined (the form required by DL.3 can always be
achieved). So in these cases the axioms are double-faced. On one hand they
are just asemantical well-formed expressions from the considered formal-
ized system, on the other hand they are meaningful propositions concerning
empirical facts and posited to be true.
Now, one of the theories of a domain of empirical (conceptual) facts,
namely deductions, is logic. Logic establishes logical laws, “tautological
propositions” that are always true exclusively in virtue of their mere form:
a composed proposition where the atomic propositions are laws of physics,
can have a form such that exclusively the truth value 1 (“true”) is assgined
to it by its truth-table, which means that the composed proposition can be
true even if some of the physical laws asserted by the atomic propositions,
or all these laws, are false. In this sense the tautological logical axioms are
closed with respect to non-logical domains of facts; they are isolated from
the truth-qualifications of the atomic propositions which concern factual do-
mains different from the logical one; they are endowed with an immutable
truth-value “true” which concerns exclusively logical empirical truth, “logi-
cal form”, being devoid of reference to any view of empirical truth different
from the view of “logical empirical truth” (if such an expression is permit-
ted). B Russell [26] wrote:

“All the propositions that are demonstrable in any admissible logical


system must share with the premises the property of being true in
virtue of their logical form; and all propositions that are true in virtue
of their logical form ought to be included in any adequate logic.”

(Here “premises” stands for “axioms”.) But a formalization of logic


can introduce also axioms that are not tautologies (the axiom of infinity,
the axiom of choice), whereby empirical truth of non-logical essence can
be also injected into a formalization of logic: this, in Russell’s view, is a
problem. Anyhow the essential point in this context is that even the logical
laws which by their tautological form express logical empirical truth, have
been constructed such by man, with the deliberate aim to codify in a per-
forming and method-offering way the domain of facts consisting of human
242 M. MUGUR-SCHÄCHTER

deductive reasoning. Logical systems are not purely formal systems, they are
formalizations of a theory which legalizes, normalizes a domain of concep-
tual facts. They build methods for the conservation and vehiculation of the
empirical truth captured in the axioms. In any acceptable formalization of
logic this basic aim entails intimate relations between logical empirical truth
and demonstrability (still a rather unexplored domain).25
25
Note added in proof. Now, what—exactly—is a “logical truth” (a tautology, a
logical axiom) according to MRC?
(2)
Consider an atomic relative proposition in the sense of DL.3, p(D, Vτ ) where D stands
(2)
for a relative description D/G, œG , V / and Vτ represents a metaview of empirical truth
supposed to exist with respect to D in the sense of D7. Furthermore, consider a metapropo-
sition composed from atomic relative propositions with the help of the usual logical op-
erators (∨, ∧, ¬, ≡). Call it a relativized composed proposition. The “logical form” of a
composed proposition is determined exclusively by the way in which the propositions
from it are connected via logical operators (and parentheses). We assert the following
(non-numbered) proposition:

ΠL. According to MRC a “logical truth”—a tautology—is a metadescription of a set


of two previously specified relative propositions (atomic or composed), with respect to
a metaview of comparison of the locations inside the global qualification space
introduced by the views V from the involved descriptions, this metadescription
yielding uniformly the result “identical location” in consequence of—exclusively—the
logical forms of the two compared propositions (so independently of the object-entities,
views and metaviews of empirical truth involved by these). Mutatis mutandis, a similar
conclusion holds concerning logical contradictions.

“Proof”. Let us begin by an example. Consider the tautology p ∨ q ≡ [(p ∧ q) ∨ (p ∧


(2)
¬1) ∨ (¬p ∧ q)]). Inside MRC-logic p and q are two relative atomic propositions p(D, Vτ )
(2)
and q(D0 , (Vτ )0 ) where D and D0 stand for two relative descriptions D/G, œG , V / and
(2) (2)
D0 /G0 , œ0G , V 0 /, while Vτ and (Vτ )0 stand for two different metaviews of empirical
truth in the sense of DL.2, supposed to exist in the sense of D7 relatively to, respectively,
D and D0 . Let us denote by cp1 and cp2 the two relativized composed propositions of
which consist, respectively, the first and second member of the tautology. Both members
involve the same qualification space, say SQ , namely that one determined by all the aspect
views involved by V or by V 0 or by both (each one counted one time). No matter what
sort of object-entity œG is considered in D, according to the definitions D14 of a relative
description D finally sends into a certain subspace or “atomic location” L(D) from SQ ,
while D0 , for any object-entity (œG )0 , sends into a certain atomic location L(D0 ) from SQ .
In the first member cp1 of the tautology these two atomic locations are then composed
accordingly to the logical form expressed there by the help of the operator ∨ of logical
sum, thus yielding a global “composed location” L1(D, D0 ) (which, according to the well-
known rules of the classical logic of propositions and its relation with the logic of classes
determined by predicates, consists in this particular case of the set theoretic sum L(D) ∪
L(D0 )).
(2)
In the second member cp2, D and D0 from the arguments of, respectively, p(D, Vτ ) and
0 (2) 0 (2)
q(D , (Vτ ) ), act like in the first member cp1, while in the arguments of ¬p(D, Vτ ) and
(2)
¬q(D0 , (Vτ )0 ), D and D0 send, respectively, into the complement of L(D) with respect to
SQ , say K[L(D)/SQ , and the complement of L(D0 ) with respect to SQ , say K[L(D0 )/SQ .
QUANTUM MECHANICS VS. RELATIVIZED CONCEPTUALIZATION 243

The preceding remarks hold also for the logico-mathematical systems.


These are constructed as formalizations of this or that domain of sponta-
neously formed intrinsic mathematical models in the sense of D19.2 (the inte-
ger numbers, the geometrical objects, etc.). By insertion into such a formal-

These 4 atomic locations L(D), L(D0 ), K[L(D)/SQ , K[L(D0 )/SQ , are then composed ac-
cordingly to the logical form of cp2, as it is determined, accordingly to the rules from
the classical logic, by the operations indicated by the symbols involved in cp2 and their
succession. This, again, sends finally into some global composed location L2(D, D0 ) from
SQ .
Consider now the identity sign “≡”. Since, once performed, cp1 and cp2 reduce to,
respectively, L1(D, D0 ) and L2(D, D0 ), this identity sign can only assert that these
two locations L1(D, D0 ) and L2(D, D0 ) from inside SQ , coincide. So, in MRC-terms,
the identity sign “≡” from the considered tautology acts like a value “identical location
(2)
inside SQ ” of a metaview VL of comparison of cp1 and cp2 with respect to an aspect
of location, say, L, inside the global qualification space SQ determined by
the views V and V 0 involved in the descriptions D and D0 from the atomic
(2) (2)
relative propositions p(D, Vτ ) and q(D0 , (Vτ )0 ) on which the tautology is
built. Furthermore, since the calculations that lead to L1(D, D0 ) and L2(D, D0 ) do
not depend on the truth valuations of p and q that can be obtained by the use of the
(2) (2) (2) (2)
metaviews Vτ and (Vτ )0 from the atomic propositions p(D, Vτ ) and q(D0 , (Vτ )0 ),
0 0
nor on the object-entities œG and (œG ) or on the views V and v from these, the
considered tautalogy asserts the identity between the locations L1(D, D0 ) and L2(D, D0 )
(2) (2)
independently of—all—œG , (œG )0 , V and V 0 , Vτ and (Vτ )0 . So this identity between
0 0
the two locations L1(D, D ) and L2(D, D ) inside the global qualification space SQ
(2) (2)
introduced by p(D, Vτ ) and q(D0 , (Vτ )0 ), emerges as a consequence of exclusively the
logical forms of cp1 and cp2.
The preceding reasoning holds in its essence for any tautology.
A similar reasoning can be constructed concerning logical contradictions (logical impos-
sibilities), and it leads to the following conclusion. The two members cp1 and cp2 of any
logical contradiction reduce to two global locations L1(D, D0 ) and L2(D, D0 ) which, with
respect to the involved global qualification space SQ , cannot both realize concerning the
unique set of atomic relative propositions from cp1 and cp2, because, in consequence of
exclusively the logical forms of cp1 and cp2, one of these locations emerges as interior to SQ
while the other one emerges as exterior to SQ , independently of the contents of the elements
(2)
œG , V, Vτ introduced by each atomic relative proposition from the contradiction.
Which establishes the asserted proposition.
Comment. So a tautology or a contradiction filters out two different “formal calcu-
lational dynamics of occupation of a location” (accordingly to the rules of the logical
calculus) relative to the one same given subspace from the global qualification space SQ
introduced by the views (in the sense of Secs. 5.1 and 5.2) from the relative descriptions
from the relative atomic propositions from the tautology or contradiction, this dynamics
being independent of the object-entities, views and metaviews of empirical truth intro-
duced by these atomic relative propositions. In a tautology these different calculational
dynamics end up with the same subspace of SQ , while in a contradiction they end up with
two mutually exclusive locations. This seems more precise a characterization than just
asserting like Wittgenstein that tautologies and contradictions are not propositions, but
“limiting cases—indeed the disintegration – of the combination of signs” (Tractatus, 4.06,
4.462, 4.466, . . . ). Instead of cryptic formulations, MRC brings forth explicit and crystal
clear ones.
244 M. MUGUR-SCHÄCHTER

ization these models are much purified, accordingly to various requirements,


and are organized in structures endowed with a strict formal coherence. The
result is endowed with a power of rigorous deductive re-expression of the es-
sential features of the initial spontaneous models wherefrom it stems, which
often is so remarkable that it is perceived as if miraculous.
But in a genuinely “pure”, non-interpreted formal system, the axioms
are not also relative propositions, they are exclusively well-formed expres-
sions from S selected as those by which a proof-description is permitted
to start: this is the specificity of an axiom from a strictly formal system,
not truth (think of formal games or of certain calculi). The axioms from
a non interpreted formal system are simply not connected with the con-
cept of empirical truth. This, however, is forgotten in the current ways of
speaking, just because formal systems which are neither interpreted, nor in-
terpretable, nor obtained from a theory of a domain of empirical facts by
axiomatization and formalization, are devoid of interest. So the double-faced
[axioms-propositions] are present in the mind as soon as one thinks of an
interesting case, and therefore it is continued to think and speak in terms
of truth of the axioms. Then, given that formal proofs start with axioms,
furthermore the intermediary well-formed expressions are often called propo-
sitions, and the theorems, having been proven, are ipso facto considered to
be also empirically true. Which amounts to a surreptitious fading away of
the case of exclusively formal characters, and a fallacious substitution to
these, of semantic-deductive characters.
Inside MRC this sort of gliding is refused by method. We are in
possession of an explicit definition of each one of the involved concepts: a
priori possibility of relative meaning in the sense of D7, piece of elaborated
relative meaning in the sense of one or the other of the definitions D14
of a relative description, relative view of empirical truth in the sense of
DL.2, relative proposition in the sense of DL.3, formal system S, view of
demonstrability inside S in the sense of DL.5, proof-description inside S. We
shall never say that the axioms from a purely formal system S are posited to
be true; nor shall we say that a theorem from S is a well-formed expression
that has been proven to be true in S, we shall only say that it has been
proven to follow from the axioms in the way required in S. And we shall
distinguish sharply between a formal system and the formalization, logical
or logico-mathematical, of a theory of a domain of physical or conceptual
facts, logic itself included.
These distinctions do by no means exclude the pertinence of the
concept of empirical truth concerning the work with “formal facts”. The
metaviews of empirical truth are here recognized to have the major role
in the intuitive pre-construction of formalizations of a theory of a domain
QUANTUM MECHANICS VS. RELATIVIZED CONCEPTUALIZATION 245

of facts—logic included—as well as in the intuitive pre-construction of com-


plex proof-descriptions. For instance, one can want to prove inside arithmetic
that given any prime number there always exists a bigger one. The above
expression of this assertion in terms of usual language can be without dif-
ficulty put in the canonical form of a relative proposition in the sense of
DL.3, defined relatively to a specified metaview of “mathematical empirical
truth” that introduces a case by case examination of truth-value consisting
in each case of the exhibition of an example. With respect to this metaview
of mathematical empirical truth one might then find that the assertion has
never been found false in any of the examined cases. This sort of empirical
(conceptual) research develops in the conceptor’s mind the preliminary in-
tuitions necessary for becoming able to attempt a proof-description inside,
say, Peano’s formalized arithmetic (how to start, what deductive trajectories
to imagine tentatively, etc.). But this preliminary empirical work is not the
researched formal proof itself, and this proof, if it can be achieved, cannot
make an explicit, declared use of the metaviews of mathematical empirical
truth that generated the intuitive knowledge of the conceptual situation:
there is no place, in a formalized mathematical proof, for metaviews of em-
pirical mathematical truth.

Gödel’s proofs versus MRC

Let us now consider the properties of completeness or decidability of a for-


malized system S (not a purely formal one), and of consistency of this sys-
tem26 . This leads to Gödel’s famous proofs [27]. These proofs establish that
(a) if Peano’s first order formalization of arithmetic, AP say, is posited to
be representable inside the formalization of logic achieved by Russell and
Whitehead [28] in Principia Matematica (PM), then AP is found not to
be complete; and (b) from this first conclusion of non-decidability of AP
it follows (with a slight generalization of P M ) that the consistency of AP
cannot be proven inside Ap either. These results hold for a large class of
other formalized systems and other formalizations of logical thought.
We shall now show that MRC throws a new light on the questions
of consistency and completeness. In the first place, it entails—quite inde-
pendently of the question of completeness—that in general the consistency
of a formal system cannot be formally examined inside this system. The
same impossibility holds concerning the completeness of the system, this
26
Completeness (or decidability) of S: the (presumed) property of S according to which
any expression that one can exhibit, which is well-formed according to S, is decidable in
S, i.e., either this expression or its negation can be proven in S. Consistency of S: the
(required) property of S according to which, for any well-formed expression from S that
can be exhibited, it is not possible to prove in S both this expression and its negation.
246 M. MUGUR-SCHÄCHTER

time quasi without reservations. So—according to MRC—examination of


both consistency and completeness, but independently of one another, re-
quire the specification of a metasystem and are then explicitly relative to
the utilized metasystem, not just properties of the studied formal system it-
self. In the second place, MRC suggests to require by method that a “good”
metasystem, offering an optimized formalization of the logico-mathematical
thinking, shall not permit inside itself undecidable expressions that can be
treated like propositions (which P M does permit). Thereby MRC displaces
the accent from a deductive point of view centered upon the studied for-
mal system, to a constructive methodological point of view concerning the
acceptable metasystems.
We begin by reproducing Sec. 1 from Gödel’s work.27

ON FORMALLY UNDECIDABLE PROPOSITIONS OF


PRINCIPIA MATEMATICA AND RELATED SYSTEMS1
by Kurt Gödel, Vienna

The development of mathematics in the direction of greater exactness


has—as is well-known—led to large tracts of it being formalized, so
that proofs can be carried out according to a few mathematical rules.
The most comprehensive formal systems yet set up are, on the one
hand, the system of Principia Matematica (PM)2 and, on the other,
the axiom system for set theory of Zermello-Fraenkel (later extended
by J. v. Neumann3 . These two systems are so extensive that all meth-
ods of proof used in mathematics today have been formalized in them,
i.e., reduced to a few axioms and rules of inference. It may therefore
be surmised that these axioms and rules of inference are also sufficient
to decide all mathematical questions which can in any way at all be
expressed formally in the systems concerned. It is shown below that
this is not the case, and that in both the systems mentioned28 there
27
Lacking the German original, the English translation, found on the web, has been
verified with the French one as published in E. Nagel, J. R. Newman, K. Gödel, and J.-Y
Girard, Le théorème de Gödel (Seuil, 1989). Taking into account both the significance of
the word and its French translation, we have substituted the word “true” to the word
“correct”, which in the English translation available to us introduced confusion. Gödel’s
notes are all reproduced—with their own numbering—after the quotation from his main
text, in order to avoid confusion with our own notes. Those among Gödel’s notes that
are irrelevant here (bibliography) are not reproduced, only their existence is indicated,
followed by dots. Our notes concerning the quotation from Gödel’s text are inserted in the
general series of our notes, but their numbers are written with Arial Black characters.
28
These are (essentially) formalizations of logic, so involving meaning and empirical
truth.
QUANTUM MECHANICS VS. RELATIVIZED CONCEPTUALIZATION 247

are in fact relatively simple problems in the theory of ordinary whole


numbers4 which cannot be decided from the axioms. This situation
is not due in some way to the special nature of the systems set up,
but holds for a very extensive class of formal systems, including, in
particular, all those arising from the addition of a finite number of
axioms to the two systems mentioned,5 provided that thereby no false
propositions29 of the type described in footnote 3 become provable.
Before going into details, we shall first indicate the main
lines of the proof, naturally without laying claim to exactness. The
formulae of a formal system—we restrict ourselves here to the system
P M —are, looked at from outside, finite series of basic signs (vari-
ables, logical constants and brackets or separation points), and it is
easy to state precisely just which series of basic signs are meaningful
formulae and which are not6,30 Proofs, from the formal standpoint,
are likewise nothing but finite series of formulae (with certain speci-
fiable characteristics). For metamathematical purposes it is of course
immaterial what objects are taken as basic signs, and we propose
to use natural numbers7 for them. Accordingly then, a formula is a
finite series of natural numbers8 , and a particular proof-schema is
a finite series of finite series of natural numbers. Metamathematical
concepts and propositions thereby become concepts and propositions
concerning natural numbers, or series of them9 ,31 and therefore at
least partially expressible in the symbols of the system P M itself.
In particular it can be shown that the concepts “formula”, proof-
schema“, “provable formula“ are definable in the system P M ,32 i.e.,
one can give10 a formula F (v) of P M —for example with one free
variable v (of the type of a series of numbers), such that F (v)—
29
Although he employs the word “false”, Gödel means here apparent propositions, not
untrue ones, as his note 4 shows: He explicitly says there that the “false” (apparent)
propositions from the metasystem are undecidable. But he continues to make use of the
word “proposition” in order to point toward these only apparent propositions. Though
in the explicit conclusion of his proof as presented in the above-quoted section 1, Gödel
did not assign a role to this fact, let us note that it was present to his mind. It will
appear below that this fact is the crucial feature for understanding the MRC-significance
of Gödel’s work.
30
Later in his proof Gödel re-expresses the meaningful sequences of signs which he
wants to make use of, in terms of defined notations that point briefly toward the logical
meaning of the considered sequence of signs (variable, proof- sequence, provable, etc.).
31
“Isomorphic” in Gödel’s note 5 means that the logico-mathematical meanings and
the truth valuations are preserved.
32
For instance, according to the PM definition no. 20, “x is an elementary formula”
is the meaning of the writing Elf (x)?(∃y, z, n)[y, z, n ≤ x & T ypn+1 (z) & x = z =
T ypn (y), and “x is a provable formula” is the meaning of the writing (in the German
original) Bew(x) = (Ey)yBx from the definition no. 46.
248 M. MUGUR-SCHÄCHTER

interpreted as to its content—states: v is a provable formula. We now


obtain an undecidable proposition of the system P M , i.e., a proposi-
tion A, for which neither A nor not-A are provable, in the following
manner.
A formula of P M with just one free variable, and that of
the type of the natural numbers (class of classes), we shall designate
a class sign. We think of the class signs as being somehow arranged
in a series11 , and denote the n-th one by R(n); and we note that the
concept “class-sign” as well as the ordering relation R are definable
in the system P M . Let α be any class-sign; by [α; n] we designate
that formula which is derived on replacing the free variable in the
class-sign α by the sign for the natural number n. The three-term
relation x = [y; x] also proves to be definable in P M . We now define
a class K of natural numbers, as follows:

n ∈ K =∼ (Bew[R(n); n])11a (1)

(where Bew x means: x is a provable formula). Since the concepts


which appear in the definiens are all definable in P M , so too is the
concept K which is constituted from them, i.e., there is a class-sign
S 12 such that the formula [S; n]—interpreted as to its content—states
that the number n belongs to K. S, being a class-sign, is identical
with some determinate R(q), i.e.,

S = R(q)

holds for some determinate natural number q. We now show that


the proposition [R(q); q]13 is undecidable in P M . For supposing the
proposition [R(q); q] were provable, it would also be true;33 but that,
on the basis of what precedes, means that q would belong to K, i.e.,
according to (1), ∼ (Bew[R(q); q]) would hold good, in contradiction
of our initial assumption. If, on the contrary, the negation of [R(q); q]
were provable, then ≡ (n ∈ K) would hold good. [R(q); q] would
thus be provable at the same time as its negation, which again is
impossible.34
The analogy between this result and Richard’s antinomy
leaps to the eye35 ; there also is a close relationship with the “liar”
33
This distinction is essential. It takes support on the logical theorem according to which
a provable universal proposition is true (Gödel’s proposition [R(q); q] is a universal).
34
PM is supposed here to be consistent.
35
According to PM, Richard’s antinomy is vitiated by the confusion between distinct
logical types in the sense of Russell’s theory of logical types, while Gödel’s proof respects
the Russellian stratification of distinct types.
QUANTUM MECHANICS VS. RELATIVIZED CONCEPTUALIZATION 249

antinomy14,36 since the undecidable proposition states precisely that


q belongs to K, i.e., according to (1), that it is not provable. We
are therefore confronted with a proposition which asserts its own
unprovability15, .37 The method of proof just exhibited can clearly be
applied to any formal system having the following features:38 firstly,
interpreted as to its content, it disposes of sufficient means of ex-
pression to define the concepts occurring in the above argument (in
particular the concept “provable formula”); secondly, every provable
formula in it is also true as regards its content.39 The exact state-
ment of the above proof, which now follows, will have among others
the task of substituting for the second of these assumptions a purely
formal and much weaker one.
From the remark that [R(q); q] asserts its own unprovability, it
follows at once that [R(q); q] is true, since [R(q); q] certainly is unprov-
able (because undecidable). So the proposition which is undecidable
in the system P M yet turns out to be decided by metamathemati-
cal considerations. The close analysis of this remarkable circumstance
leads to surprising results concerning proofs of consistency of formal
systems, which are dealt with in more detail in Sec. 4 (Proposition
XI).

36
Again the connection (“false proposition, i.e., antinomy)-(undecidability) on which
our note 29 draws attention.
37
Gödel’s note 15 is remarkably curious. It concerns exclusively the process of construc-
tion (“projection” in Peano’s arithmetic) of the (“false”) “proposition” [R(q); q], and of
identification of the meaning imparted to it by this formal construction, while the original
content, inside PM, of this “proposition”, is not criticized. The process of construction of
[R(q);q] indeed is not circular, it respects Russell’s requirement of stratification of the log-
ical types, etc.. But the proposition itself, by its original content, is “antinomic”. Gödel’s
note 14 and his own expression “false propositions” (to which our note 29 refers) testify
that he was fully aware of this and that he researched precisely such an antinomic struc-
ture, in order to be able, by taking support on it, to reject its decidability in terms of
empirical truth and therefrom to infer also an undecidability in terms of a formal proof
inside Peano’s arithmetic. But, eventhough he starts by announcing a general critical atti-
tude with respect to the metasystem PM, it remains cryptic in his subsequent formulations
whether, specifically, he considered acceptable the possibility of “antinomic propositions”
inside PM; and correlatively, whether he considered to have indicated a way for con-
structing better metasystems than PM, or to have definitively established that Peano’s
arithmetic is not decidable, if it is consistent (as it seems to be involved by the current
ways of speaking).
38
What follows in the text shows that “formal system” means here a formalization of
logic, or more generally a metamathematical system intended to be able to include and
to rule mathematical systems or questions. It does not mean Peano’s arithmetic.
39
See our note 32.
250 M. MUGUR-SCHÄCHTER

Notes
1. . . .
2. . . .
3. . . .
4. I.e., more precisely, there are undecidable propositions in which,
besides the logical constants ≡ (not), ∆ (or),) (x) (for all) and =
(identical with), there are no other concepts beyond + (addition)
and · (multiplication), both referred to natural numbers, and where
the prefixes (x) can also be referred only to natural numbers.
5. In this connection, only those axioms in P M are counted as
distinct as do not arise from each other only by change of type.
6. Here and in what follows we shall always understand the term
“formula of P M ” to mean a formula written without abbreviations
(i.e., without definitions). Definitions serve only to abridge the
written text and are therefore in principle superfluous.
7. I.e., we map the basic signs in one-one fashion on the natural
numbers (as actually done on p. 179)
8. I.e, a covering of a section of the number series by natural
numbers. (Numbers cannot in fact be put in a spatial order).
9. In other words, the above-described procedure provides an
isomorphic image of the system P M in the domain of arithmetic,
and all metamathematical arguments can equally well be conducted
in this isomorphic image. This occurs in the following outline
proof, i.e., “formula”, “proposition”, “variable”, etc., are always
to be understood as the corresponding objects in the isomorphic
image.
10. It would be very simple (though laborious) actually to write out
this formula.
11. Perhaps according to the increasing sums of their terms and, for
equal sums, in alphabetical order.
11a The bar-sign indicates negation (replaced with ∼).
12. Again there is not the slightest difficulty in actually writing out
the formula S.
13. Note that “[R(q); q]” (or—what comes to the same thing “[S; q]”—
is merely a metamathematical description of the undecidable
proposition. But as soon as one has ascertained the formula S,
one can naturally also determine the number q, and thereby
effectively write out the undecidable proposition itself.
14. Every epistemological antinomy can likewise be used for a similar
undecidability proof.
15. In spite of appearances, there is nothing circular about such a
proposition, since it begins by asserting the unprovability of a
wholly determinate formula (namely the q-th in the alphabetical
arrangement with a definite substitution) and only subsequently
(and in some way by accident) does it emerge that this formula
is precisely that by which the proposition was itself expressed.”

Let us comment on this inside MRC (see, as an introduction, the


final global comment on the definitions D14 of a relative description). For
the sake of clarity we continue to proceed by sequences proposition-“Proof”.
We begin by an assertion related with the last paragraph from the above
quotation, concerning consistency.
QUANTUM MECHANICS VS. RELATIVIZED CONCEPTUALIZATION 251

ΠL.7. Proposition on consistency. According to M RC the ques-


tion of the consistency of a formal system S cannot, in general, be settled
inside S, for reasons that are independent of any assumption concerning the
completeness of S. In general this question can be settled only by formal
examination inside a conveniently constructed metasystem M S. Then the
solution established inside M S is relative to M S.

“Proof”. The consistency of S is by definition the (required) prop-


erty of S according to which, for any well-formed expression from S that
can be exhibited, it is not possible to prove inside S both this well-formed
expression and its negation (note 26).

Now, the whole qualificational power defined inside S, is concen-


trated in the view of demonstrability VdS from the proof-descriptions
DjS /GSj ,œj ,VdS /. In each one of these the object-entity œj consists by def-
inition of just one well-formed expression from S: VdS does not exist in
the sense of D7 with respect to “any well-formed expression from S that
can be exhibited”—a potential meta-entity with respect to those, œj , from
the achieved proof-descriptions DjS /GSj ,œj ,VdS /— so VdS cannot qualify this
meta-object-entity as a whole. However, though S says nothing concern-
ing the way in which one may “exhibit” well-formed expressions different
from those enumerated ab initio in the definition of S, otherwise than by
achieving proof-descriptions DjS , it might happen that somehow—with the
help of projections from some metasystem M S, or by empirical research—a
well-formed expression from S be first found without any proof DjS , that
can then be proven inside S, via the view of demonstrability VdS , as well
as its negation. Thereby the in-consistency of S would be proven inside S,
by construction, and the question would be closed. But this is a particu-
lar circumstance which may stay indefinitely non realized; and as long as
a proof of inconsistency by construction has not been produced, the ques-
tion of the consistency of S stays open. Or otherwise, in the case of certain
trivial finite systems S, it can be possible to produce one by one all the
well-formed expressions permitted by S, and to study them by correspond-
ing proof-descriptions DjS /GSj ,œj ,VdS /, thus concluding inside S concerning
the consistency of S. But in general an assertion of consistency of S cannot
be founded on a sequential production of well-formed expressions from S. In
general such a process is not efficient because there is no way to ascertain
that the production is finished, nor that, while it continues, inconsistency
will never be found, nor that, if for a given well-formed expression œj no
proof of inconsistency is found, none is possible. So according to M RC the
question of the consistency of S cannot—in general—be settled inside S. It
252 M. MUGUR-SCHÄCHTER

follows that only a formal examination of S as a whole, achieved from the


outside of S, could settle this question.
But this, according to M RC, requires another sort of description than
the proof-descriptions DjS /GSj ,œj ,VdS / from S, where not [S-as-a-whole] is
the object-entity. Indeed the principle of separation P15 asserts that “Since
any one relative description D/G,œG ,V /, whatever its complexity, involves
by construction one generator of object-entity, one object-entity, and one
view, all well defined, as soon as some change is introduced in the content or
the role designated by a term from the triad G,œG ,V , another description is
considered”. And, by method, P15 posits that “this other description must
be treated separately”.
Now, since a formal proof is researched, it must be achieved inside
some formal system, namely some convenient metasystem M S inside which
S be somehow embeddable.
Suppose then that such a metasystem has been found and that inside
it a proof of the consistency or the inconsistency of S has been achieved.
Then nothing excludes that with another metasystem (M S)0 6= M S the con-
clusion of this proof be contradicted: Though inconsistency can in principle
happen to be provable inside S by an example—i.e., in an absolute way—in
general a proof concerning the consistency of S is relative to some metasys-
tem M S.
So πL.7 is entirely established.

Comment. The fact that in general a proof of consistency of a


formal system, requires a metasystem, is well known. The new element here
is only that (and how) this follows inside MRC, and quite independently of
considerations concerning the completeness of the studied system.

We consider now Gödel’s proof of undecidability.

ΠL.8. Proposition about the expression [R(q); q]. According to


MRC the well-formed expression [R(q); q], by construction, is not a propo-
sition, so it cannot be true or false. So Gödel’s reductio becomes impossible
and aimless.

“Proof”. According to MRC, a relative description D/G ,œG ,V /is a


piece of elaborated meaning where the three roles G, œG , and V have all to
be defined, and played accordingly to their definitions, by definite epistemic
(2)
actors.40 Furthermore a relative proposition p(D, Vτ )(DL.3) involves a def-
40
Even if in a degenerate way (i.e., two roles are held by one same actor) and/or without
radicality (the generator G does not radically create the object- entity œG , the view V does
not radically change this object-entity while qualifying it) (cf. the final general comment
QUANTUM MECHANICS VS. RELATIVIZED CONCEPTUALIZATION 253

inite relative description D/G,œG ,V / that has been previously established


independently, and afterward is subjected to valuation by the truth-values
(2)
of some definite metaview of empirical truth Vτ that exists with respect
to D/G,œG ,V /, in the sense of D7 (DL.3).
The metasystem P M dwells with well-formed expressions that can
be “interpreted as to their contents” (meanings) and with respect to these
can be a priori awaited to be found to be empirically true or false. So,
though implicitly, descriptions and propositions are involved in P M . Then,
according to MRC, what is the descriptional status of the formula [R(q); q]
from P M ?
[R(q); q] is not a relative description D/G,œG ,V /. Indeed [R(q); q] is
first constructed by a succession of syntactical steps. Once obtained in this
way, it is “interpreted as to its content” (cf. Gödel’s text) and found to
assert its own unprovability. But [R(q); q] consists exclusively of this self-
qualifying assertion. The roles of generator of object-entity and of object-
entity are not defined, so they are not played. In “I am not provable” of
what does “I” consist? [R(q); q] asserts the unprovability of nothing defi-
nite. The fact that [R(q); q] has been constructed in full agreement with all
the syntactical requirements from P M (the stratification of distinct types
included), does not change this situation.
Now, inside MRC the definition D14 of a relative description (cf.
the final global comment) banishes explicitly self-referential constructs, on
semantic-methodological grounds, whatever their grammatical or logico-
mathematical correctness. So according to MRC there is no description cor-
responding to [R(q); q]. A fortiori [R(q); q] is not a relative proposition in
the sense of DL.3 either.41 It is not even a proposition in the loose sense of
the classical logic, (i.e., [(an assertion) (that can be true or false)]), since the
first element, an assertion, in its own right, is lacking), its place is held by a
sequence of signs which, though it is correct from a syntactical viewpoint, is
devoid of any possible meaning by its very inner structure. As Gödel himself
says, it is a false (apparent) proposition (cf. the end of the first paragraph
from our quotation of Gödel’s text, and our note 37 on Gödel’s note 15). In
these conditions [R(q); q] is doomed not to be provable in P M : there is no
way to prove the truth of something that is not a proposition.

of the definitions D14).


41
In order for R(q); q] to be a proposition in the sense of DL.3 it would have been
necessary to first specify inside PM a well-formed expression, say XP M , which, interpreted
as to its content, be a definite description D/G,œG ,V /; and then, in order to “propose”
tentatively that the meaning (the content) carried by XP M is true, to construct [R(q); q]
as a genuine (universal) proposition, i.e., such that, considered itself now as to its content,
(2)/ (2)
it be found to be a metadescription D(2) /G2) , œ(2) , Vτ with œ(2)≡D/G,œG ,V / and Vτ
a metaview of empirical (logico-mathematical) truth.
254 M. MUGUR-SCHÄCHTER

So Gödel’s reductio becomes impossible (and aimless). The hypothesis


“supposing the proposition [R(q); q] were provable, it would also be true”
is known a priori to be impossible, by construction, which dissolves the
reductio.

Comment. In so far that one is aware, as it does happen inside


MRC, that a linguistic construct like [R(q); q] does not exist in the sense
(2)
of D7 with respect to any view Vτ of empirical truth (be it logical or
mathematical empirical truth), Gödel’s reductio is settled in advance and
acquires the character of a game of play pretend.42

ΠL9.On the MRC-significance of Gödel’s proof. According to


MRC, the main previously unknown result established by Gödel’s proof is
that the metasystem P M permits inside it well-formed expressions that are
not decidable, so are not propositions, and that these can be injected into
Peano’s arithmetic AP by isomorphic projection. There, via examinations
monitored by P M , they reproduce their non decidability.

“Proof”. Obvious from Gödel’s proof and the preceding “Proof”.

Comment. Strictly expressed, according to MRC Gödel’s proof es-


tablishes a conclusion about the metasystem P M + AP , not about Peano’s
arithmetic AP considered independently of P M . The undecidability proven
by Gödel is relative to the Russell-Whitehead metasystem P M . So the
propositions πL.8 and ΠL.9 displace the accent from the studied formal sys-
tem, on the metasystem which is made use of for the study. But thereby one
is led to a further quite general question, analogous to the question concern-
ing consistency examined in πL.7: is it conceivable to study the completeness
of a formal system S from inside S?

ΠL.10. Proposition on completeness. According to MRC, in gen-


eral the question of the completeness of a formal system S cannot be settled
inside the system. It requires the use of a convenient metasystem MS. This
entails that the result is relative to MS.

“Proof”. Completeness of a formal system S is the (presumed)


property of S according to which any expression that one can exhibit, which
is well-formed according to S, is decidable in S, i.e., either this expression
42
J-Y. Girard, in Le champ du signe ou la faillite du réductionnisme, in Le
théorme de Gödel, by E. Nagel, R. Newman, K. Gödel, and J-Y. Girard (Seuil, 1989),
writes: “Si l’on dégage les idées profondément novatrices—essentiellement la distinction
vrai/démontrable—autour desquelles se charpente le théorme, la démonstration résulte
d’une suite impitoyable de truismes—ou de ‘provismes.’”
QUANTUM MECHANICS VS. RELATIVIZED CONCEPTUALIZATION 255

or its negation can be proven in S.


The argument from the “proof” of πL.7 can be transposed in an
obvious way.

Comment. In these conditions, speaking of “the” completeness of


S as if it were an absolute property of S, is in general misleading. In general
the property of completeness of a formal system is radically dependent on
the metasystem that is made use of for establishing its existence.
This leads us to ask whether the features of a metasystem M S, which
permit to induce in a studied formal system S, undecidable well-formed
expressions, are indeed unavoidable features. The statement πL.8 seems to
indicate a negative answer. Indeed, since MRC—a nonformalized method—
does avoid the emergence of “undecidable false propositions”, a fortiori it
should be possible to build also a formalization of logic which avoids such
emergences (as well as any other features that can generate undecidability).

The pertinent question, in this respect, seems to consist of the speci-


fication of methodological rules for constructing “good” metasystems.

The above (very rapid and quasi informal) examination of the ques-
tions of consistency and completeness of a formal system illustrates well the
fundamental difference between classical logic and MRC. In classical logic
all the creative epistemological features are occulted by storage in the ab-
solutized and hypostatized concepts of “values of an object-variable x” and
of a shadow-predicate “P ”. The involved descriptional relativities are not
apparent, hence their consequences also remain hidden, so they cannot avoid
false problems, nor show the way toward the natural solution when prob-
lems do creep in. Whereas inside MRC the double-extremity genetic classes
expose explicitly all the involved cognitive actions, so the relativizing conse-
quences of these upon the produced qualifications are obvious. Furthermore,
the limitations entailed by the descriptional relativities are explicitly tied to
a methodological obligation P15 to interrupt the current descriptional pro-
cess and to take a new start on a metalevel, which organizes in cells the
conceptual progression and keeps it under control.

5.1.3. Conclusion on the MRC-logic

It is remarkable that MRC, such as it has been constructed by taking initial


support exclusively on quantum mechanics, leads to the outline of a logical
approach that is relevant not only for the basic, the physical creative genetic
classes C[G(o) .V (o) ] of the type of those involved in quantum mechanics, but
also for conceptual creative genetic classes C[G.V ] found to be involved in
formal systems.
256 M. MUGUR-SCHÄCHTER

The quantum mechanical cognitive strategy, generalized inside MRC,


has opened up a way of conceptualization that is not mute with re-
spect to the most fundamental questions of nowadays abstract math-
ematical and logical thinking.

This is so because the canonical descriptional mould (G, œG , V )


drawn from quantum mechanics has been constructed at the lowest level
of conceptualization which human mind has been able to reach, possibly the
final one. There the most severe conditions that can be encountered in a
process of conceptualization, are all active. So a basic structure of labelled
receptacles for conceptualization which is constructed to fit these conditions,
is sufficiently comprehensive for harbouring any descriptional possibility that
might occur. Inside this structure, semantics and cognitive actions—which
always involve aims —combine with the syntactic features, and this induces
both intelligibility and control.

5.2. MRC versus Probabilities


One of the major successes of MRC is the representation of a deeper general
concept of probability, which contains and explains the so cryptic quantum
mechanical probabilities [14,16,17A,18]. Indeed, when Kolmogorov’s classi-
cal concept of a probability space is examined inside MRC, the limitations
and the absolutizations which flaw this concept come into striking evidence.
By suppressing them, the concept of probability expands to the limits of
its whole natural volume which rests on the most basic level of transferred
conceptualization and extends up to very high descriptional levels.
Throughout the process of construction of the MRC-concept of proba-
bility, the methodological principle of separation P15 plays a key role. There-
fore this process can also be regarded as a succession of illustrations of the
very peculiar way in which the principle of separation works.

5.2.1. Komogorov’s classical definition of a probability space

The fundamental concept of the modern theory of probabilities—in Kol-


mogorov’s formulation [29]—is a probability space [U, τ, p(τ )] where: U =
{ei } (with i ∈ I and I an index set) is a universe of elementary events ei
(a set) generated by the repetition of an “identically” reproducible procedure
P(called also an experiment) which, notwithstanding the posited identity
between all its realizations, nevertheless brings forth elementary events ei
that vary in general from one realization of P to another one; τ is an algebra
of events built on U ,43 an event, let us denote it e, being a subset of U
43
An algebra built on a set S is a set of subsets of S—S itself and ∅ being always
included—which is such that if it contains the subsets A and B, then it also contains
QUANTUM MECHANICS VS. RELATIVIZED CONCEPTUALIZATION 257

and being posited to have occurred each time that any elementary event ei
from e has occurred; p(τ ) is a probability measure defined on the algebra of
events τ .44 A pair [P, U ] containing an identically reproducible procedure P
and the corresponding universe of elementary events U is called a random
phenomenon.
On a given universe U , one can define various algebras τ of events.
So it is possible to form different associations [[random phenomenon],[a cor-
responding probability space]], all stemming from the same pair [P, U ].
With respect to the previous representations (Bernoulli, von Mises,
etc.)—where only a concept of “probability law” (or “probability measure”)
was defined mathematically—Kolmogorov’s concept of a probability space
[U, τ, p(τ )] has marked a huge complexifying progress.

5.2.2. Critical remarks

In Kolmogorov’s classical theory of probabilities, the procedure P is neither


formally defined, nor symbolized or otherwise represented. This theory con-
tains no symbolic location reserved for the procedure P, so a fortiori the
random phenomenon [P, U ] as a whole is not represented. The consequence
is that the structure of the connection between the considered probability
space [U, τ, p(τ )], with the substratum wherefrom it is generated, is very
rarely explicitly surveyed. Usually nothing whatever is asserted concerning
the way in which the elementary events from the universe U do emerge by
the procedure P.

The channel for the adduction of semantic substance from the “pool
of reality” (in the sense of D2) into the considered probability space
[U, τ, p(τ )], is undefined and unexplored. It is only alluded to by mere
words.

In each application of the abstract theory of probabilities, to some


specific problem, the corresponding semantic substance is injected into the
studied probability spaces in an intuitive unruly way. It might be argued
that this is an intentional non-determination which endows the formalism
A ∪ B and A − B.
44
A probability measure defined on τ consists of a set of real numbers p(A), each one
associated to an event A from τ , such that: 0 = p(A) = 1, p(U ) = 1 (normation), p(∅) = 0,
and p(A∪B) ≤ p(A)+p(B) where the equality obtains iff A and B are “independent” in the
sense of probabilities, i.e., iff they have no elementary event ei in common (A ∩ B = ∅).
The number p(A) yields the value of the limit – supposed to exist—toward which the
relative frequency n(A)/N converges when the number N of realizations of the involved
repeatable procedure P is increase